Beruflich Dokumente
Kultur Dokumente
-1-
Urban roads
Urban major road junctions
Urban major road links
Urban town centres
Urban Residential roads
Rural roads
Rural Villages
Motorways
Rural Dual-carriageways
Rural Dual carriageway junctions
Rural Dual carriageway links
Rural Major single-carriageways
Rural Major single-carriageway links
Rural Minor single-carriageways 42
Rural Minor single-carriageways junctions
Rural Minor single-carriageways links
Non-engineering measures
5. Measuring effectiveness
Monitoring matters
When to monitor
Control sites
Overall assessment of monitoring
Data quality
Police involvement and publicity
Accident monitoring
Monitoring vehicle speeds
Automatic speed monitoring
Radar or laser speed monitoring
Where to monitor vehicle speeds
Journey time monitoring
Flow monitoring
Monitoring pedestrian movements
Monitoring pedal cycle and two-wheeled motor vehicle movements
Monitoring vehicle (time or distance) headways 50 or inter-vehicle gaps
Monitoring traffic conflicts
Attitude surveys
Road user surveys
Environmental monitoring
Evaluation
Evaluation of traffic speed data
Evaluation of public perception
Evaluation of accident changes
Standard tests of accident changes
-2-
Acknowledgements
The substance of this Guide was prepared by TRL Ltd. The main authors at TRL were Judith Barker and
Chris Baguley. They would like to acknowledge the many staff at TRL who made significant
contributions to both drafting parts of the Guide and advising on its content.
The authors are also indebted to all those highway authorities, central government departments and other
organisations who responded either to a questionnaire or direct contact requesting information about their
recent good examples of accident remedial schemes and their views on the contents of the Guide. The
authors are particularly grateful to the following authorities who willingly provided information and
photographs on specific sites:
Bracknell Forest Council
Ceredigion County Council
Glasgow City Council
Hertfordshire County Council
Kingston-upon-Hull City Council
London Borough of Sutton
Luton Borough Council
Oxfordshire County Council
Norfolk County Council
Southampton City Council
South Tyneside Metropolitan Borough Council
Sunderland City Council
Swindon Borough Council
Vale of Glamorgan Council
Wolverhampton City Council
York City Council
Most of the photographs in the main body of the text are provided courtesy of TRL Limited.
1. Introduction
-3-
Purpose
1.1 This is the second publication in the DTLRs good practice advice series on local transport. This guide
shares good practice, to help achieve the targets set out in DTLRs road safety strategy Tomorrows roads
safer for everyone (DTLR, 2000b).
1.2 This guide has been developed primarily as a reference for local authority staff with an interest in road
safety engineering and associated issues. However, we also hope that it will interest a wider audience,
including the police, the Highways Agency, local health authorities, local communities, businesses, and
transport interest groups.
1.3 It is hoped that the guide will particularly benefit those new to road safety engineering, whether they
are just starting their careers or are transferring from a related discipline.
1.4 This guide is intended to be a living document, which will be updated over time as knowledge and
experience develop. Consequently other existing and new examples of good practice will be sought and
we particularly welcome feedback from practitioners (see chapter 8).
Scope
1.5 A number of documents that advise on various aspects of road safety 1 management, including the
design of engineering measures and schemes, are already in the public domain. This document aims to
draw together existing advice as far as possible into one document, and to update it, based on the most
recent experience of local authorities and agencies (see examples, Appendix A), and on research results.
1.6 Consequently, the level of detail varies, and it follows that this guide is not intended to be a fully
comprehensive document to be used in isolation. Where a subject is covered in depth in up-to-date
publications elsewhere, the subject may be dealt with more briefly and reference will be made to the other
sources of information in the text. Additional references that may be of use are given only in the
bibliography. Where there is little current published information available, we deal with the topic more
fully in this guide.
1.7 The good practice in this guide is not only based on the full Local Transport Plans (LTPs) submitted
by local authorities, but also on other aspects that DTLR considers good practice. Also included are
innovative examples provided by local authorities, thought to be successful, but which are too recent to be
proven so.
1.8 Strenuous efforts have been made to provide accurate, up-to-date and full coverage of the issues
relating to road safety engineering and good practice, with the focus on engineering. However, it should
be noted that much of the contents is taken on good faith and some subjective judgements have been relied
upon in the choice of approach. Readers should also be aware that what constitutes good practice in one
authority, or on a particular road, or under one set of specific conditions, may not be good practice if
simply replicated elsewhere. This is due to the complexity of the many interacting factors affecting safety.
Similarly, it should also be noted that there is not one definition of good practice, as no single definition
could cover everything.
-4-
Structure
1.9 This guide has eight chapters:
Chapter 1 comprises the above purpose and scope of this guide and a background introduction to
road accidents in Great Britain. It also describes how safety problems are currently being tackled;
Chapter 2 looks at the management of road safety on the network, including staffing, training,
planning, liaison and consultation processes. This chapter also considers the roles of road safety
officers, road safety engineers and safety qualifications;
Chapter 3 describes the general principles of road safety work. The role of Local Transport Plans
(LTPs) is discussed. It provides methods of identifying and prioritising problems using accident and
casualty analysis techniques (including the need to consider urban and rural problems separately and
how to take exposure to risk into account particularly for vulnerable road users). The chapter also
deals with finding solutions to accident problems, prioritising schemes, and economic justification in
terms of accident reductions. This chapter also raises issues relating to the funding, installation,
Safety Audit, and monitoring of schemes;
Chapter 4 (in conjunction with Appendix A) comprises the main body of the guide. It describes
specific national safety problems(including accident and casualty statistics not published
elsewhere)and a selection of engineering measures offered as potential solutions, according to
location and road type;
Chapter 5 describes many of the methods available to monitor and evaluate the success of schemes,
including some which may be useful when accident numbers are small and the levels of exposure to
risk are unknown;
Chapter 6 is the bibliography. It contains:
the contents list for the Traffic Signs Manual (TSM chapters one to eight);
a list of Traffic Advisory Leaflets (TALs);
a list of Local Transport Notes (LTNs);
a list of the highway standards and advice notes found in the Design Manual for Roads and Bridges
(DMRB); and
a large list of other publications relevant to road safety, including those referred to in this guide. The latter
are listed alphabetically by author;
Chapter 7 lists abbreviations used in this guide;
Chapter 8 describes how readers can give feedback to DTLR;
Appendix A contains brief descriptions of individual engineering measures and key references on
proven performance, where available. The appendix also gives one or more example
schemes(generally) submitted by local authorities, with a brief description of their purpose and
performance;
Appendix B gives basic details on applying various statistical techniques, including some worked
examples;
Appendix C is the standard data input form for the safety scheme accident monitoring database,
MOLASSES.
-5-
Background
1.10 Over the last 50 years, the number of motor vehicles in Great Britain has increased dramatically.
Motor traffic levels are about ten times greater today. In 1950 there were about 550 people injured in road
accidents every day, about 14 of who were fatally injured.
1.11 Both central and local government have made substantial efforts to reduce the road accident toll and,
despite the large increase in traffic, accident rates (per vehicle-km travelled) have declined dramatically.
In1966, the number of fatalities peaked at 22 per day. Since then the number of fatalities has gradually
fallen. In 1999, there were about nine and a half per day.
1.12 Many factors are involved. However, we can identify three major contributions to the drop in
casualty numbers and their severity:
a) legislation requiring that seat belts be fitted (1967) and worn (1983onwards);
b) drink-driving legislation (the current legal limit of 80mg of alcohol in100ml of blood was introduced in
1967) and changes in the attitude of the public to drink-driving;
c) traffic calming schemes on local roads (1980s onwards).
1.13 In addition, in 1987 the government of the day focussed attention on road safety by setting a national
target to reduce road accident casualties by one third by the year 2000 (compared with the
1981-85average). A great deal of effort and initiatives followed and since that target was set, road injuries
have fallen by only 0.5 per cent but road deaths have fallen by 39 per cent and serious injuries by 45 per
cent. However, the volume of traffic has increased by 160 per cent over the same period, and those
slightly injured may previously have been seriously injured or killed.
1.14 Compared with the rest of the world, and the rest of Europe in particular, we have one of the lowest
levels of road deaths per head of population, and per licensed motor vehicle (DTLR, 2000a). We also have
one of the lowest car occupant deaths per car-km (Automobile Association AA, 1999a).
1.15 However, 320,310 people were still injured in a single year in road accidents in Great Britain (1999
data) of these casualties 3,423 were fatally injured nearly nine and a half a day (DTLR, 2000a) 2 .
Although in many respects our roads are safe compared with others in Europe, our child pedestrian
fatality rate is one of the highest in Europe (DTLR, 2000b).
1.16 While most road accident casualties in Great Britain are car occupants, this reflects the fact that the
car is the main type of road transport.
Other types of road users are subject to a higher risk of accidents, despite their relatively low exposure on
the network (in terms of time spent or distance travelled). These are known as vulnerable road users and
include pedestrians, pedal cyclists, motor cyclists and equestrians. Many of these road users are children
who suffer serious injuries, as they are generally less well protected in accidents than people in cars. The
safety needs of vulnerable road users are extremely important.
-6-
1.17 DTLR published a national cycling strategy in 1996 (DTLR, 1996). Encouraging walking: advice to
local authorities was published in March 2000 (DTLR, 2000c). It is a working guide for the people who
turn policy into action. It is not an exhortation to the public to walk more, but aims to make it safer, easier
and more convenient for them to choose to do so. It shows how improvements to the walking environment
can be made at both the strategic and the tactical level.
1.18 Accidents place a large financial burden on the nation, not only in terms of the costs associated with
personal injury but also in terms of damage to property. The current estimated annual savings from
preventing all (including about 3,520 million damage only) road accidents is 16,310 million (DTLR,
2000a).
1.19 These statistics reflect the need for a continued, concentrated and well-managed approach in order to
reduce the number and severity of accidents.
1.20 DTLR has published a road safety strategy Tomorrows roads safer for everyone (DTLR, 2000b)
The strategy sets casualty reduction targets for the year 2010 (with progress to be measured by
comparison with 1994-98 averages) 4 . The 2010 road safety casualty reduction targets are:
a 40 per cent reduction in the number of people killed or seriously injured in road accidents 5 ;
a 50 per cent reduction in the number of children (under 16years of age) killed or seriously
injured; and
a 10 per cent reduction in the slight casualty rate, expressed as the number of people slightly
injured per 100 million vehicle kilometres.
1.21 The road safety strategy also outlines strategies to tackle problems associated with:
children and the most vulnerable road users;
driver training;
drink, drugs and drowsiness;
the road infrastructure;
speeds;
vehicle design and maintenance; and
enforcement, education and information.
1.22 Alongside the safety strategy, DTLR published New Directions in Speed Management A review of
policy (DTLR, 2000d). This contains the latest information on the relationship between vehicle speed and
accidents and other factors including air quality, noise, quality of life, and health.
1.23 Directly or indirectly, everyone can influence road safety. We are all responsible for controlling the
risk we expose ourselves to and the risk we subject others to. No one is excluded from using the road
network, so everyone must work towards the common goal of accident reduction. Stakeholders will
include:
individual road users using the network for work or leisure (motorists, pedestrians, motorcyclists,
pedal cyclists and equestrians, including those with special needs, such as those with disabilities);
and
those in national and local government, Northern Ireland, the Scottish Executive, the National
-7-
Assembly for Wales, government agencies, Transport for London, local authorities, the police,
health, education, public transport, commercial companies, charities, research bodies and special interest
groups.
1.24 There have been changes recently in the way local highway authorities(LHAs) bid for a share of
DTLR funding. The Department has replaced the old Transport Policies and Programme (TPP) system
with Local Transport Plans (LTPs) 6 . The funding for safety under TPP was ring-fenced and distributed
annually, based to some extent on the road safety plans produced by local authorities. From 1999-00,
authorities submit LTPs that cover all aspects of local transport management, including safety. Capital
funding for road safety measures is now awarded as part of a wider block of local transport funding over
which local authorities have discretion. From 2001-02, DTLR has given an indication of local transport
funding levels for five years.
1.25 Two recently published documents give guidance on developing full LTPs (DTLR, 2000e) and
examples of good practice in LTP development (DTLR, 2000f).
1.26 The first of these documents states that LTPs should describe the specific road safety policies in a
local road safety strategy and that the strategy must contain local casualty reduction targets for
2005,broken down into annual milestone targets, so that local authorities can monitor progress.
Authorities will want to establish targets to reflect the national targets.
1.27 The document goes on to outline the expected contents of a local safety strategy, including details of
proposed and existing engineering schemes and non-engineering approaches. It emphasises the need for
monitoring and the benefits of contributing to Great Britains central database, known as MOLASSES
(Monitoring Of Local Authority Safety SchemES see Appendix C and paragraph 5.4 below). It requires
the production of an annual progress report and tables of performance indicators for measuring progress.
Guidance has been produced to assist authorities in monitoring and reporting on progress in implementing
their Local Transport Plans, including progress towards all local objectives and targets (DTLR2001).
1.28 If safety is to improve, local highway authorities must allocate resources for staff and materials
within the overall programme. These resources should be sufficient to design for safety, and allow
authorities to implement, monitor and assess good practice. In the long term, if a coherent approach is
adopted, society ought to profit from such an approach in terms of accident savings and an improved
quality of life.
1
This guide is concerned with road safety only. It does not deal directly with the security of persons or
property.
2
These include the figures for motorway and trunk road accidents, which comprise 12 percent of all
casualties and, because of relatively high severities on trunk roads, 18 per cent ofall fatalities.
3
The equivalent for motorway and trunk roads is the Highways Agencys Safety Strategy, safety being
one of eight strands of an overall strategy. The strategy aims to spread good practice and help those
without any safety training. The Highways Agency published it in March 2000 (Highways Agency,
2000a).
-8-
These targets are also those adopted by the Scottish Executive and the National Assembly for Wales.
Implementation of the strategy will be taken forward by the UK and devolved administrations in
accordance with their respective roles and responsibilities. Northern Ireland will have its own road safety
strategy.
5
The target for the motorway and trunk road network is a reduction of 33 per cent (Highways Agency,
2000a).
6
Note that separate systems are in place in London, Wales and Scotland.
-9-
2.9 Councillors (elected members), parish councils and local committees represent local communities who
need to be kept informed of road safety issues and planning. Public consultation is invaluable and you can
find practical advice in, for example, IHT (1996a) and IHT (1999b).
2.10 The structure and organisation of traffic policing varies from area to area. However, in general,
traffic police will assist road users, enforce road traffic law and supervise any necessary temporary
changes in traffic management. The police, ambulance and fire services may all be present at the scene of
an accident. The police also play a significant role in encouraging, publicising and educating road users
about traffic law. Sometimes the police will also be involved in Safety Audit and accident investigation.
2.11 The Crime and Disorder Act (1998) requires police and local authorities to carry out a three yearly
audit of problems affecting their area and draw up a plan to deal with them (DTLR, 2000b). This should
include road safety.
2.12 Therefore, the police 9 need to be informed of any relevant changes to the network, such as scheme
installation, special events and road works. The links between the police and local authorities are
invaluable and, indeed, they often collaborate in strategies such as speed management and setting speed
limits, the provision and operation of speed cameras, training programmes, driver rehabilitation projects.
In addition, the decriminalisation of parking offences and the proposed decriminalisation of unauthorised
use of bus lanes will raise road safety issues and opportunities. The police can also advise on road traffic
patterns and driver behaviour because of their specialised knowledge of the network.
2.13 Perhaps most importantly, the police collect all the information on reported road traffic accidents.
For each accident resulting in personal injury, the police record the circumstances surrounding the
accident and details of the vehicles and casualties. They may also include contributory factors, participant
and witness statements, photographs or sketch plans. The accident, vehicle and casualty information is
included in the national STATS19 accident database and this helps to identify the problems that road
safety strategies must tackle.
2.14 Consultation between LHAs and HA agents is also important, particularly on schemes where the
trunk road network adjoins or crosses the local network and for joint publicity initiatives. In addition, the
trunk road network has been divided into core and non-core 10 elements and in the near future local
authorities will take over responsibility for the non-core elements (DTLR, 1998a). It will take excellent
liaison for the handover to go smoothly.
2.15 The various departments within a local authority should liaise to avoid conflicts of interest and to
provide a cost-effective service. There is particular value in incorporating the efforts of RSOs and RSEs in
the planning stages and especially when introducing innovative schemes. We generally recommend that
communication with colleagues in:
policy;
planning and development;
maintenance;
highways (engineers responsible for new roads and major schemes);
public transport;
traffic management and control;
access and mobility; and
- 10 -
trading standards.
2.16 Consultation and co-operation between local authorities and outside bodies is also important. For
example, develop links with:
the general public;
local residents and businesses;
landowners and farmers;
motoring organisations;
driving instructors;
cycle and motorcycle trainers;
HGV driving instructors;
freight transport groups;
special interest road user groups;
public transport companies;
construction companies;
charities;
professional organisations (e.g. IHT, ICE, IRSO etc);
environmental groups;
private companies;
teachers;
local health authorities; and
hospital trusts and other organisations (such as CAPT, RoSPA, CPRE etc).
2.17 There is often potential for joint projects perhaps targeting a road safety problem from more than one
direction to make the most of financial and human resources. Some examples of this approach include:
local highway engineers working with the local health authority on cycle scheme initiatives, to help
meet both road safety and health of the nation targets;
a local authority and a car manufacturer jointly funding training schemes for young drivers.
2.18 Local authorities can also benefit from co-working, sharing information and experience and acting
consistently:
locally, working with other authorities through, for example, the regional Government Offices and
accident reduction working groups, Local Authority Road Safety Officers Association
(LARSOA),Association of London Borough Road Safety Officers (ALBRSO),Transport for London
and so on;
at a national level through, for example, the DTLR, Highways Agency, County Surveyors Society,
Parliamentary Advisory Committee on Transport Safety (PACTS), Local Government Association
etc.
2.19 Some guidance on forming partnerships is given for the purpose of developing LTPs (DTLR, 2000e
and 2000f).
- 11 -
2.20 Disseminating the results of local authority road safety work to a wide audience is essential. This can
be done through formal or informal presentation and through publication, either individually or by
gathering information from several sources together. DTLR regularly gathers, analyses and interprets this
sort of information to develop its advice see also chapter 6.
2.21 It may be a good idea to set up a formal approach to consultation and liaison on a regular basis. The
Gloucester Safer City project found this useful (DTLR, 2001b). For example, meetings could be
pre-arranged at regular intervals and an annual diary of events, campaigns and initiatives could be issued.
- 12 -
2.30 An engineering scheme immediately affects all road users who travel through it and its effects are
also measurable.
2.31 Chapter 3 outlines the general principles for road safety engineering work. Chapter 4 details a
number of safety engineering accident-remedial treatment measures for use on a wide range of roads.
Chapter 5 deals with the measurement and evaluation of the effects of safety engineering schemes.
Appendix A contains examples of successful schemes installed by local authorities across Great Britain.
2.32 Road safety schemes must be maintained, just like the road network generally. Liaison with those
responsible for maintenance (eg operating the UK Pavement Management System) is essential to ensure
that new schemes are integrated within the maintenance programme. For optimal safety of the network,
we generally recommend that you maintain the best value indicators specified as minimum levels for
principal roads (BVPI96 see DTLR, 2000j) and for non-principal roads (BVPI97).
Safety on the motorway and trunk road network is the responsibility of the agents acting onbehalf of the
Highways Agency.
- 13 -
10
11
A few personnel have qualifications that are no longer offered: Road Safety Engineering MSc
(Middlesex University) is a notable example.
12
SafeNET is used to estimate the frequency of accidents on an urban network (from traffic and
pedestrian flow and geometric information provided). OSCADY, ARCADY and PICADY are used as
aids to the design of signalised, roundabout and priority junctions, respectively.
3. General principles
3.1 This chapter outlines the general principles associated with road safety work, and engineering in
particular. It begins by describing the role of road safety in Local Transport Plans (LTPs). It then goes on
to describe accident and casualty analysis techniques to identify road safety problems. The next sections
address how to find solutions to the problems and to prioritise work programmes and issues relating to the
installation and Safety Audit of safety schemes. The final section stresses the importance of learning from
the success or failure of a scheme.
- 14 -
- 15 -
3.12 It is important that authorities should monitor the progress they are making in implementing their
Local Transport Plans. Authorities have been asked to prepare Annual Progress Reports which identify
the progress being made in working towards the local objectives, targets and outputs contained in their
Local Transport Plans, including all road safety targets. These reports will also help to indicate the link
between local and national targets, and may help to identify any barriers to the achievement of local
objectives or targets.
- 16 -
- 17 -
Transport for London, London Accident Analysis Unit (LAAU)publishes regular reports on the
accident and casualty data for all the London authorities (see London Research Centre, 1999a,
1999b,1999c for examples);
Publications by interest groups, such as the Child Accident Prevention Trust (CAPT, 2000a).
3.26 To explore b) and c) in paragraph 3.24 above, the possible safety effects of encouraging more cycling
and walking (DTLR, 1996; DTLR, 2000c) and the possible effects of the Road Traffic Reduction
Act(1997 and 1998) should be considered. Some useful publications are:
Transport Statistics Great Britain. This is an annual publication (DTLR, 2000h). DTLR also produce
quarterly summary information;
National Travel Survey; and
National Household Census.
3.27 Factors affecting d) in paragraph 3.24 above may include almost anything from the installation of a
scheme (within the last three to five years), to temporary road closures, to a large sporting event, to a
petrol shortage, to a new law or publicity campaign.
3.28 Any changes to the local network in terms of the road length understudy should also be taken into
account. For example, the introduction of lower speed limits in villages and the new responsibilities for
non-core trunk roads may both affect the proportion of rural and urban roads (and so casualties) in the
network year by year.
3.29 It may be helpful to tabulate the results of the analysis, to assess the relative seriousness of problems
to help prioritise them. This will be particularly useful in identifying any overall problems, such as
speeding, skidding, or bend accidents. It may provide justification for a mass action treatment (see
paragraph 3.17 above).
3.30 The information used on accident rates during the prioritisation process will vary, depending on the
situation and the quantity and reliability of exposure data. The use of more than one type of accident
analysis approach will often be appropriate.
3.31 When ranking problems a balanced assessment of all the data has to be achieved, based on:
accident rate (see paragraphs 3.41 to 3.44 below);
number of accidents; and
severity of injuries sustained in accidents.
3.32 An intervention level is a numerical value of a measure of an accident problem (such as
accidents/year, accidents/vehicle-km). If the values for a particular road exceed the relevant intervention
levels, then select that road for more detailed analysis and subsequent treatment.
3.33 Over time, most of the worst accident problem sites have been cured. Accidents now tend to be
spread more evenly across whole areas. For this reason, mass action, route action or area action remedial
treatments may be preferable to treatments at a few specific sites (see paragraph 3.17 above). The
treatments selected may be chosen to tackle one or more particular types of accident, rather than all
accidents.
- 18 -
3.34 Low-cost measures may make these other approaches just as cost-effective as the traditional
site-specific approach. In addition, some accident problems may be tackled more effectively through
enforcement, training and publicity than by engineering alone.
Data issues
3.35 The nationally collected accident database using material from theSTATS19 form, described in
paragraph 2.13 above contains objective information about accidents. A sample form can be found in the
back of RAGB (DTLR, 2000a). STATS20 (DTLR, 2000g) accompanies theSTATS19 form. It advises on
the meaning of certain aspects of the form. The following points may also help in accident analysis:
Many villages have 30 or 40 mph speed limits 22 . They are therefore classified as urban, even if the
surrounding roads are rural and the land is predominantly not built on. It is not easy to identify such
sites from accident data, except with the use of maps.
The road type (was carriageway type) variable in STATS19 does not include the categories
roundabout and one way street as either single or dual carriageways. The two together account for
approximately 10 per cent of all accidents so take care to account for them in analyses that consider
single and dual carriagewaysseparately. 23
There is no specific definition of a bend or the severity of a bend.
This is because in terms of safety many factors are important, including the type of approach, camber,
super elevation, radius, transition, road surface, aspect, verge width, gradient etc. It is up to the reporting
officer to decide how to classify the accident. In addition, a bend feature is only specified as a category in
the vehicle manoeuvre variable. This means that the reporting officer may record the vehicles in an
accident that happens at a junction on a bend as either a junction or a bend-related manoeuvre.
Horses are now included as a vehicle type on the STATS19 form 24 .
Police authorities should record all injury accidents. Of course, not all accidents are reported to the
police. Research using hospital and insurance records has shown that recording levels for accidents
involving vulnerable road users and for those involving only a single vehicle are particularly low (eg
Mills, 1989; James, 1991).
There is some evidence that the precise location given for an accident is often inaccurate. It may be
hard to ascertain precisely where the accident occurred, with respect to where the vehicles came to a
halt, particularly in the case of a high speed accident. Also, the police will not always attend the
accident scene immediately after the accident.
- 19 -
3.37 Fig. 3.1 shows the trend in the proportion of casualties on rural roads compared with the 1981-85
averages, by severity.
Fig 3.1: Indices of the proportion of all casualties that were on rural roads(including motorways),
by injury severity (1986-99)
3.38 Rural roads (excluding those through villages) are less likely than urban ones to be treated with
safety engineering schemes. The main reasons for this are that:
local authorities tend to identify sites for treatment on the grounds of injury accident numbers; and
accidents in rural areas are even more likely than those in urban areas to be widely scattered.
- 20 -
3.39 Intervention levels (see paragraph 3.32) suitable for urban (but not rural) roads will be well known to
local authority RSEs from their knowledge of their local network. Intervention levels suitable for use in
rural accident analysis 25 (including some relating to vulnerable road users) have been developed
recently, based on national data (Barker et al, 1999). A methodology for rural accident analysis using
these intervention levels is given in DTLR (2001a) and Barker et al (2001).
3.40 There is also another good practical and financial reason for considering the separate analysis of
urban and rural roads. The Highway Maintenance Code of Good Practice (LAA, 1989)recommends
creating a highway maintenance management strategy and a maintenance road hierarchy of urban and
rural roads separately. It recommends that it should be further broken down by traffic flow and
composition. Efficiencies will result if maintenance and safety scheme programmes can work together as
suggested in paragraph3.74 below.
Importance of exposure
3.41 It is well known that accidents are highly correlated with traffic flows and road length (for example,
see Walmsley and Summersgill, 1998). In other words, you would generally expect to find more accidents
on a long road with high flows than on a short road with low flows. It is important, therefore, to take
account of exposure (or the opportunity for accidents to occur) when ranking accident problems.
3.42 Although it is also important to tackle the largest number of accidents and casualties possible with
the budgets available, it is important to note that this does not necessarily mean treating the sites with
most accidents. This is because sites with a high accident risk (ie the sites that do not have the largest
number of accidents, but do show a greater propensity for accidents than one would expect for a given
level of exposure to risk) are the sites that are most likely to be amenable to treatment.
3.43 Although the most important exposure variables are likely to be road length and vehicle flow, others
will often be important too, especially when considering certain types of accidents. Examples include
pedestrian flow, pedal cycle flow, junction and bend density (the number of junctions or bends per km of
road).
3.44 In the case of vulnerable road users this approach is especially important as, although accident and
casualty numbers may be low, their accident casualty risk is very high.
- 21 -
3.48 This method of comparing accident types between areas identifies accident types that feature
disproportionately more often than their expected frequencies 27 would suggest.
3.49 To illustrate, if only 5 per cent of all accidents in the study area involve pedal cycles then, in terms of
accident numbers, it is not immediately apparent that this is a high priority accident problem. Now,
suppose we find that only 1 per cent of all accidents in a similar control area involve pedal cycles.
Assuming that the level of cycling is similar in the two areas, it is evident that cycle accidents in the study
area deserve further investigation. It is possible that such an accident type (because of its abnormally high
level of risk) will be more amenable to preventative engineering treatments than another accident type
with a higher observed and expected frequency of accidents.
3.50 Statistical tests may be used to evaluate whether the difference between the proportions of each
accident type are unlikely to be due to chance (ie the difference between 5 per cent and 1 per cent in the
example above). The statistical test used will depend on whether the control area is mutually exclusive of
the study area see Appendix B5for details.
3.51 The use of the control does not, of course, prove that a problem exists. No control is perfect and the
results may reflect different levels of exposure between the study and control areas. The control data
could equally well camouflage a particular problem. Nevertheless, the use of a control is a valuable way
of identifying accident characteristics for further investigation.
- 22 -
Fig 3.3: Accidents rates (per 108 vehicle-kms)by road class, GB 1999
Finding solutions
Detailed investigation
3.54 The next step in tackling the road safety problem is to investigate further the sites 28 identified as
having problems. Once this has been done, and solutions identified a cost-benefit analysis (see 3.77
below, for example) will provide the final information required to prioritise remedial treatments.
- 23 -
3.55 It is important to consider road hierarchies. The actual use of different parts of the network should be
assessed as it may not be in keeping with the design function of the road. If changes need to be made to
address accident problems, then the method and overall potential effect of such changes will need to be
addressed.
3.56 A more detailed accident and casualty analysis of the possible remedial sites will normally include
plotting accidents on a larger scale map and considering the commonality of accident types and the
movements of vehicles and pedestrians.
3.57 Stick diagrams 29 can be useful to identify predominant accident types, but become unmanageable if
the number of accidents is greater than20 or so.
Site visits
3.58 Site visits are an essential part of the detailed investigation process. Ideally, they should involve
more than one person visiting a site on more than one occasion, in different weather, lighting and traffic
conditions.
3.59 On-site observations of layout, signing, markings and traffic movements can often reveal possible
explanations as to why accidents are happening that are not apparent from studying maps and accident
reports alone.
3.60 The main reference sources for road layout, signing and marking are in paragraph 3.68 to 3.71 below
and are listed in full in the first part of Chapter 6, the bibliography.
3.61 Typically, a site inspection should involve a road safety engineer addressing the following types of
questions.
Is there a consistency and clarity of approach in the quantity, quality, type and standard of
maintenance of layout, signs and markings along the road?
Is correct warning of a hazard given on the approach by use of the proper road markings hazard
centre line, SLOW marking and so on as set out in the Traffic Signs Manual (TSM), Chapter 5 (see
bibliography)?
Are the markings and road studs properly maintained, so that they are clearly visible by day and by
night and give the necessary minimum preview time (two seconds)?
Are the prescribed warning signs provided (see TSM Chapter 4 for guidance)?
Are the signs the correct distance from the hazard?
Are the signs the correct size for the prevailing traffic speeds?
Are the signs properly maintained and in good condition? (Worn or dirty signs cannot deliver the
intended level of service.)
Can the signs be clearly seen over the full recommended visibility distance or are they obscured by
foliage, other signs, parked vehicles etc?
Are the signs sited under trees or otherwise in deep shadow for much of the day? If so they may be
prone to algae growth which obscures them and seriously degrades retroreflective performance.
Fluorescent backing boards will greatly enhance conspicuity in daytime, when ultraviolet radiation is
present.
Are signs difficult to see because they are viewed against a complex background? (If so a yellow
- 24 -
Detailed design
3.64 It is important that the objectives of introducing a scheme (such as speed reduction, improving
pedestrian safety and so forth) are clearly thought out before beginning the detailed design process.
3.65 An outline scheme design should be drawn up which may include several different approaches and
engineering measures for achieving the objectives of the scheme. The next stage is to identify the
individual elements of a scheme and to put them together to form a cohesive, detailed design.
3.66 Engineers should consider all aspects of safety throughout the design process, as specified in the
Contract (Design and Management)Regulations 1994 (HMSO, 1994a).
3.67 Engineers should consider the needs of all road users, including the disabled, pedestrians and
motorists. They should also consider social inclusion 30 and accessibility.
3.68 The Traffic Signs, Regulations and General Directions (TSRGD HMSO, 1994c) provides drawings
of all prescribed Traffic Signs together with relevant Regulations and Directions that apply to them.
Detailed information and advice concerning design and application of signs, markings and other
engineering measures are given in:
- 25 -
Traffic Signs Manual Chapters one to eight (DTLR, various dates).The Chapter titles are given in the
front of the bibliography and are referred to in the text as TSM Chapter No.
Traffic Advisory Leaflets (DTLR, various dates). These are listed herein the front of the bibliography
(Chapter 6) and are referred to in the text as TAL No./yr.
Local Transport Notes (DTLR, various dates). Relevant ones are listed in the bibliography (after the
Traffic Advisory Leaflets). They are referred to in the text as LTN No./yr.
Design Bulletin 32 (DTLR, 1992) and its companion guide (DTLR,1998c) cover the design of
residential road and foot way layouts.
Transport in the Urban Environment (IHT, 1997) promotes the design and management of urban
transport infrastructure and systems. Part III includes safety related issues.
3.69 The principal documentation for trunk roads is:
the set of Design Standards and advice notes brought together in Design Manual for Roads and
Bridges (DMRB DTLR, various dates). The constituent parts of this document are given here in the
bibliography (after the TALs and LTNs, Chapter 6) and referred to in the text as DMRB, XX No./yr..
Some of DMRB may be appropriate when considering non-trunk roads but the possible effect of
differences in traffic flows and traffic mix should be reviewed.
3.70 If non-prescribed signs or markings are to be used then non-prescribed sign authorisation must be
sought from DTLR.
3.71 The current advice on speed limits can be found in Circular Roads1/93 (DTLR, 1993) and TAL 1/95.
Since changes in legislation in 1999(DTLR, 1999c; Scottish Office, 1999; Welsh Office 1999), local
authorities are free to introduce self-enforcing 20 mph zones and 20mph speed limits, where appropriate.
3.72 The design process will include drafting clear, well-annotated, vertical and horizontal drawings for
the engineers. Outline plans should be on a scale of 1:5000 for a route. Otherwise they should be 1:2500or
1:1250. Full detailed plans should be on a scale of at least 1:500, or1:200 for a more complex scheme.
3.73 Consultation will largely take place after an outline scheme design has been proposed and before the
full detailed design is finalised. Consultation can take up a significant amount of time and budget during
the design process. Those to be consulted will vary from scheme to scheme, but will usually include
internal consultees, local residents, emergency services and other representatives or road user groups
likely to be affected by the scheme (see Chapter 2).
3.74 Ideally, consultations will include establishing that no other works are planned in the near future for
the road that is to be treated, preferably for at least a three year period. This will ensure that:
the scheme has a reasonable future lifespan;
scheme monitoring will not be compromised; and
different work schedules may be combined (such as maintenance and scheme installation) so that
some necessary costs will be incurred only once.
3.75 It is important that sufficient money is set aside for monitoring a scheme. In the long term,
monitoring will justify future similar schemes and make their design and installation more cost-effective.
- 26 -
3.76 The various stages of scheme design, consultation etc should all be well documented to reduce the
amount of work necessary should a similar scheme be installed in future.
Economic justification
3.77 The economic justification for installing a safety scheme is usually based on its economic return.
This is generally calculated as an estimated First year rate of return (FYRR) which is an estimate of the
monetary benefits to be gained in accident savings in the first year set against the cost of the scheme 31 .
While many schemes will only save a small number of accidents a year, this can still produce a good rate
of return.
3.78 Sometimes, particularly for the large schemes, the lifetime of the scheme may be taken into account
with a net present value being calculated (calculated in a similar way to compound interest Highways
Agency, 1996).
3.79 The estimated average accident prevention savings for accidents and casualties for 1999 (DTLR,
2000a) are:
The estimated average accident prevention savings for accidents and casualties for 1999 (DTLR,
2000a)
3.80 Predicting the percentage of accidents and casualties that are expected to be saved by a scheme is a
difficult task. Consideration needs to be given to the recent accident history, the type of accident the
scheme is designed to minimise, and the effect of the scheme on other (and potentially new) types of
accident. The accidents to be saved should relate directly to the type and objectives of the scheme.
3.81 In some instances, robust information on the performance of particular measures may be available as
a result of previous monitoring. For example, try the MOLASSES database, the SafeNET software
(TAL08/99; TRL Limited, 1999), Traffic calming in practice (CSS et al,1994a) and as detailed in
- 27 -
- 28 -
Using this approach the impact against each of these objectives should be recorded in an Appraisal
Summary Table. This table presents an unbiased summary of the impacts against the Governments
objectives. To be consistent with NATA the summary should be supported by an assessment of how the
measure meets local objectives and the extent to which problems are addressed. NATA also includes
analyses of the impacts on distribution and equity, practicality and public acceptability, and affordability
and financial sustainability.
The required level of detail will be proportionate to the size of the scheme. The general principle is that
the appraisal should be conducted at a level of detail sufficient for the value for money of the proposal to
be demonstrated clearly. A working note Appraisal of LTP: advice on simplified procedure provides
guidance on the level of detail that may be appropriate for different policy instruments. For the majority
of smaller safety schemes the key requirement is to ascertain that accident savings are likely and that
together with any other associated benefits they exceed total costs imposed.
Installation issues
3.90 For urban roads, Transport in the Urban Environment (IHT,1997) deals with many relevant issues.
The main documentation for scheme installation on trunk roads is the Manual of Contract Documents for
Highway Works (DTLR, 1998b). However, much of this is relevant for works on local (particularly major)
roads and local authorities often use it.
3.91 Qualified personnel who understand safe working procedures must supervise the installation of
engineering schemes. There are legal issues to be considered.
- 29 -
3.92 Full and dated records should be kept detailing each phase of installation. This site diary information
is often lacking, particularly when the client hands over aspects of the design and installation to one or
more outside bodies. However, it is crucial for the monitoring process to provide valid results.
3.93 There is a need to consider any safety issues that may arise specifically during the installation period.
This is particularly relevant when installation is spread over several days, or even months, and if the site is
to be left unattended.
Safety Audit
3.94 Safety Audit is a procedure introduced to prevent accidents. It is not an accident reduction procedure.
In Safety Audit, safety (or accident potential) is formally, and objectively, considered at each stage of the
design of a scheme.
3.95 A full Safety Audit will have four stages (IHT, 1996b):
stage 1 feasibility/initial design;
stage 2 preliminary design/draft plans;
stage 3 detailed design; and
stage 4 pre-opening (as soon as practical after completion).
3.96 A team of safety auditors carries out Safety Audits. They should be personnel with safety expertise
and who are independent of the design team. The Safety Audit team will comprise more than one person
and will include road safety engineers. It might also include police officers, road safety officers and other
specialists, such as structural engineers. The size, breadth of expertise and number of members of the
team will depend on the size and character of the scheme.
3.97 A prime objective of a Safety Audit is to consider the safety of all types of road users under all types
of conditions, such as weather and time of day. The Safety Audit will not only consider the scheme itself,
but its potential impact on the surrounding network. A necessary part of all Safety Audits will be to
balance the needs of different road users (including those using all motorised transport, vulnerable road
users and the disabled) under different conditions by assessing levels of accident risk. Naturally, the audit
will also have to consider financial and design constraints.
3.98 Considering safety throughout the design process minimises the likelihood of unforeseen factors
affecting the final costs. A final audit before opening the scheme is essential. This is because some aspects
may be difficult to consider from two-dimensional plans and because installations may not match plans
precisely.
3.99 In the longer term, Safety Audits encourage good design. They give safety a higher profile in the
design process and act as a conduit for informing engineers of current safety understanding. The
recommendations of safety auditors are not based on checking individual design elements against
standards, but on considering how the scheme as a whole may affect overall safety, or deciding what to do
when standards conflict.
- 30 -
3.100 It is recommended that a policy is adopted for Safety Audits to be carried out for all new road
schemes and all modifications to the existing network, including re-design and maintenance.
3.101 Nationally, Safety Audits are only mandatory for trunk road schemes. The level of local authority
safety auditing varies widely. Some authorities audit all schemes fully. Others reduce or eliminate the
number of Safety Audit stages, depending on the scheme size, type or cost. The amount of documentation
and the procedures for arbitration in the case of safety/design conflicts of interest also differ greatly
between authorities. It is important that each authority has a clearly defined strategy and procedure for
Safety Audit. The IHT (1996b)suggest that Safety Audit is a part of a broader road safety strategy:
priorities for audit need to be set within the total programme of highway schemes.
3.102 There is considerable guidance on the legal consequences of Safety Audit (eg Stewart, 1995; Heath,
1995). Note that legal actions might result following accidents at a scheme, particularly if the adopted
procedures have not been followed and are not well-documented.
3.103 When highway works are commissioned to outside bodies the need for Safety Audits should be
identified and documented when the contract is let. The various stages should be monitored and well
documented throughout the scheme planning, design and implementation. The processes and
responsibilities for departures from procedures and recommendations should be clearly specified.
3.104 There is now a wealth of information, advice and experience on Safety Audit. Several sources are
listed in the bibliography, but useful examples include:
Guidelines for Safety Audit of Highways (IHT, 1996b)
Standards for Road Safety Audits (DMRB, HD 19/94) and Advice
Note for Road Safety Audits (DMRB, HA 42/94)
What goes wrong in highway design (AA, 1999b).
3.105 In addition, some authorities and other practitioners have set up forums and support groups where
safety auditors can discuss common problems and solutions.
3.106 Some authorities advocate Safety Audits of existing roads. These maybe useful to give a consistent
approach or message to road users. This is particularly the case across networks where individual
elements have been designed separately over time. However, many issues maybe identified during the
detailed design stages of accident analysis and reduction programmes.
- 31 -
3.108 Do not repeat good schemes if better ones are available and appropriate.
3.109 Some schemes work well in some situations but will be inappropriate in others.
3.110 Some measures may work better alone or may only be effective when used in combination with
other measures.
3.111 Sometimes an unsuccessful scheme need not be rejected outright. Investigations may prove that it
only needs minor modifications to turn into a success.
3.112 Sometimes measures designed to reduce speeds may not do so, but may still reduce accidents by
increasing driver awareness.
3.113 Sometimes it is hard to accept failure for financial or political reasons. However, it is important to
be courageous and acknowledge that things did not go to plan. Consider challenging established policy
and procedures if the evidence from monitoring warrants it.
3.114 Sometimes the effects of one scheme are hard to assess, perhaps because accident numbers are low
or other factors influence the results. These problems can often be minimised or eliminated by considering
data for a large number of similar schemes together. For example, pooling data related to changes at each
site to reach an overall average figure. Such research can provide robust evidence about a scheme design
in a way that is not feasible at a local level.
3.115 The MOLASSES database contains basic information about schemes installed on local authority
roads and on trunk roads. The database is populated and managed at TRL (see web site at
www.trl.co.uk/molasses). Contributors can interrogate the database for the latest evidence on the overall
effectiveness of a type of measure in terms of accident reduction. It is planned that regular reports will
also be produced. See Appendix C for an example data entry form.
3.116 Over the years, DTLR and HA have funded a wealth of research into the safety effects of
engineering measures and safety schemes. Some have been studied only in the test and development
stages using off-road trials and simulator studies. The most promising ideas have gone on to be installed at
sites across Great Britain and further monitoring undertaken to establish the effects of these on-road trials.
The results of these research projects are usually published or provoke further research. They are used in
the development of the engineering design Standards and Advice notes, Traffic Advisory Leaflets,
software such as SafeNET (TAL 08/99; TRL, 1999), and to shape road safety policy.
3.117 The TRL library has a vast amount of published research and maintains a database of international
road research publications. A charge is made to provide some of this information.
3.118 Many of the more relevant publications are given in the bibliography, even if they have not been
referred to directly in the text.
3.119 Chapter 4 describes accident problems at a national level and discusses potential engineering
solutions (text in shaded boxes), by road type. It also includes a short section on non-engineering
measures.
- 32 -
13
Note that separate systems are in place in London, Wales and Scotland.
14
For more details on the New Approach to Appraisal (NATA) see DTLR 2000e and paragraph 3.87.
15
Urban (or built-up) roads are defined as those with speed limits of 40 mph or less.
16
Rural (or non-built-up) roads are defined as those with speed limits of 50 mph or more. Roads through
villages with speed limits of 20, 30 or 40 mph are not included. A rural road may or may not have
buildings alongside it.
17
Accident risk is a general term for the likelihood of an accident occurring, given a certain level of
exposure.
18
If accident numbers are high (hundreds or thousands) then one years data may be sufficient. However,
if numbers are small and the data are broken down further into small groups by type of accident, for
example, then the data will vary too much between years or sites for meaningful comparisons to be made
and may be misleading. Much more than three to five years data will lead to a tendency for changes in
flow and significant changes in the network to affect the accident picture.
19
The analysis of the types of accident and the causal factors contributing to the accident is a vital step to
reach an understanding of why accidents occur and how to treat the problem. Some of the most important
aspects to be studied include casualty severity, weather and road surface condition, road layout and
junction type, vehicle manoeuvres, vehicle types, vehicle speeds, driver compliance with the Highway
Code, driver age, pedestrian involvement etc.
20
This should include an analysis of types relevant to local and national targets and performance
indicators in LTPs ie will include accidents involving child casualties, vulnerable road user casualties and
analyses by severity.
21
See also DTLR (2001a) and Barker et al (2001) for detailed methods.
22
Government policy is now that speed limits in all villages should be 30 mph.
23
The convention adopted in the casualty analyses in this guide is to combine these data with
dual-carriageway data for motorway and A-road accidents and to combine them with single-carriageway
data for accidents on lower class roads.
24
In 1999, 181 horse rider casualties were reported, two of whom were fatally injured. 40 percent of
casualties occurred on urban roads and 60 per cent on rural roads (DTLR, 2000a).It is not known how
many horses were injured or involved in road accidents where the rider was uninjured.
25
Intervention levels (see paragraph 3.32) suitable for motorway and trunk road accident analysis are
provided annually to route managers under the Highways Agency Safety Strategy.
26
Very often national data are suitable and readily available in RAGB (DTLR, 2000a). Other suitable
Control areas might include the rest of the county, a group of neighbouring counties etc.
- 33 -
27
28
Hereon, the term site may refer to short or long sections of road, or to a network of roads.
29
A stick diagram is a table with one column for each accident. Each row of the table represents an
accident, vehicle or casualty category (eg junction type, number of vehicles, type of vehicle, casualty
severity etc). Pictograms showing vehicle movements or other categories are often used to facilitate quick
analysis. The columns in stick diagrams are sorted by one or more row categories, repeatedly, so that
predominant category types become apparent.
30
Including issues that may affect road users travel patterns such as personal safety at night, the ability to
combine parenthood and work etc.
31
The simplest FYRR will be estimated as 100*(number of accidents in 12 months before installation
predicted number of accidents in 12 months after installation)*(average cost of an accident)/total scheme
costs. Note that there will be a considerable degree of uncertainty associated with the estimation of any
such economic indicators.
4. Treatment selection
Engineering measures
4.1 The nature of road safety problems commonly varies according to type of location. Due to the
differing land use and traffic mix on the UKs roads, appropriate solutions will also vary. In the following
sections, we look at common problems and safety engineering treatments for improvement under a simple
location classification of URBAN 32 and RURAL 33 , and the road types within these, as shown in the
tree chart below.
Figure 4.1
- 34 -
Figure 4.2a
- 35 -
Figure 4.2b
- 36 -
4.2 Fig 4.2a above for reported road accidents in 1999 from STATS19 data clearly shows that the
majority of casualties (over 68 per cent) occur on urban roads. Fig 4.2b shows the predominance of urban
roads is slightly less marked when killed and seriously injured (KSI) casualties only are considered, with
59.1 per cent on urban roads. This reflects the generally higher severity of accidents on rural roads.
However, achieving the Governments casualty reduction targets still means that all road authorities need
to investigate safety problems and come up with solutions.
4.3 As discussed in Chapter 2 and in paragraph 3.8 above, the national casualty reduction targets need to
be disaggregated to annual figures for the individual authoritys road network. The authority will need to
decide which combination of strategies of single site, mass action, a reaction or route treatment is likely to
best achieve these targets. The remainder of this chapter discusses, in general terms, the main safety
problems found at different types of road location in Great Britain. It also suggests common solutions.
More details on each type of treatment together with real examples are in Appendix A.
Urban roads
Figure 4.3a
Figure 4.3b
All fatalities on urban roads 1999
- 37 -
4.4 Urban areas are more complex than rural ones and are where the majority of casualties occur (ie
built-up roads in DTLR, 2000a). These figures mask large differences between the figures for different
road users.
4.5 Fig 4.3a shows the breakdown of casualties on urban roads by casualty type. About 19 per cent are
pedestrians, 10 per cent are cyclists, 9 per cent are two-wheeled motor vehicle users and the remainder (63
per cent) are users of other motorised vehicles. When fatalities alone are considered (Fig 4.3b) the figures
are 47 per cent (pedestrians), 7 per cent (cyclists), 15 per cent (two-wheeled motor vehicle users) and 32
per cent (other motorised vehicle users).
4.6 The types of road user involved also differ substantially from one location to another. In town centres,
casualties are often concentrated at specific locations. Outside these areas, they are more diffuse and
include a markedly higher proportion of pedestrians and cyclists, particularly children (IHT, 1997).
4.7 The principle technique for tackling urban accidents is black spot treatment. Low cost solutions are
applied to clusters of accidents with a factor in common. This approach has met with considerable
success, but is becoming less beneficial as more and more sites are treated.
4.8 Urban Safety Management (USM) principles (TAL 3/90; IHT, 1990)were developed to address the
urban accident problem more strategically, tackling both clustered and more thinly spread accidents. The
approach was first demonstrated in five towns in the Urban Safety Project (Mackie et al, 1990). By
managing traffic onto the right roads, the project achieved casualty reductions of 15 per cent.
4.9 The key elements of USM are:
- 38 -
- 39 -
4.17 The number of casualties in accidents at junctions on major roads in the urban areas of Great Britain
is currently 68,876 (1999 data). This represents 22 per cent of the total casualties on the network. The
large volume of road users passing through them on conflicting paths often worsens the serious safety
problems at these junctions. For the same level of turning traffic flows, an uncontrolled priority junction
with a minor road will usually have more accidents per year than other junction types.
4.18 The speed of vehicles approaching a junction and the possibility of overtaking manoeuvres on the
major road will also directly affect the number and severity of collisions. At traffic signals the risk of
serious right angle collisions is increased if drivers infringe the red light, and the incidence of these
infringements normally increases when drivers are approaching at high speeds despite the use of
sophisticated speed discrimination or assessment systems (Baguley and Ray, 1989). These systems help
those drivers who, by virtue of their position, may be genuinely caught in a dilemma if a green signal
were to change, by extending the green phase. However, the systems only work up until the time at which
the pre-set maximum green time is reached.
4.19 Important considerations are thus to incorporate, either in the design of junctions or as remedial
action, features that will help to ensure slower speeds through the junction and increased awareness of
drivers, for example:
effective signing (DTLR, 1994);
central refuges (Appendix A11);
vehicle-activated signs (Appendix A25);
other visual cues;
speed cameras (Appendix A4);
red-light cameras (Appendix A3); and
- 40 -
- 41 -
4.29 Roundabouts have the advantage of slowing vehicle approach speeds on all arms of the junction, but
are generally less safe for two-wheelers. All roundabout design and modifications should now cater for
increased safety of cyclists and motorised two-wheelers. Continental design, with tighter entries, can help
these road users (TAL 9/97; Davies et al, 1997).
4.30 For roads with heavy traffic flows the only solution may be to physically separate pedestrians by
means of footbridges or subways. There is, however, a need to take account of pedestrians general
reluctance or inability to take longer routes or apparently unnecessary steps or slopes, and concern about
their own security (eg fear of underground passages and crime). Successful grade separation keeps the
pedestrians on the level following their desired path, whilst vehicles undergo the change in level.
Important issues such as siting, sight line, lighting, dimensions etc need careful consideration (IHT, 1997;
DTLR,1998c and HA, 1996 for Trunk Roads contains useful design guidance).
4.31 On major road links in towns and cities in Great Britain, 24,428 people were injured in road
accidents in 1999. This is about 8 per cent of all reported casualties.
4.32 Pedestrians tend to minimise their walking journey and will cross major roads where it is convenient
to do so and not always where it is safest. The safest policy is normally to minimise conflict points
between vehicles and pedestrians so that driver attention can be focussed at designated controlled crossing
places.
4.33 Cyclists comprise 7 per cent of all casualties and motorised two-wheeler riders 8 per cent. In these
accidents it is often the case that larger vehicle drivers fail to notice two-wheelers, probably due to their
smaller physical size. Indeed, motorcycle and pedestrian accidents can be a particular problem in
congested areas. Although studies of bicycle accidents have shown that most collisions with cyclists
involve turning manoeuvres at junctions, more than a third are non-junction accidents, with cyclists often
being hit from the rear. Accidents involving both motorcyclists and pedestrians are also a problem,
particularly in congested conditions and at traffic signals.
- 42 -
4.34 On high flow major urban roads, where physical speed control devices are inappropriate to help
reduce accidents, methods of automatic speed enforcement or speeding notification are in use to deliver
accident reduction. These include speed cameras (Appendix A4) 34 and vehicle-activated speed signs
(Appendix A25).
4.35 Adequate crossing facilities for pedestrians, with provision for disabled pedestrians, need to be
provided along links. These include:
refuges (Appendix A11);
zebra crossings (LTN 1/95, 2/95; DTLR, 1998d);
school crossing patrol (RoSPA, 1990; Appendix A7);
signal controlled crossings (pelican crossings; puffin crossings which incorporate pedestrian
detection; toucan crossings which incorporate a cycle crossing facility (TAL 10/93; LTN 1/95; LTN
2/95); pegasus crossings which cater for equestrians, cyclists and pedestrians); and
grade separated crossings (ie subways or footbridges IHT, 1997).
4.36 Guard rail or fencing to channel pedestrians to the designated crossing may be deemed necessary on
busy roads. However, their use should only be considered where the risks of walking onto the carriageway
are very high, as they have a number of disadvantages. They are visually intrusive, reduce foot way width,
can obscure children, and can cause access difficulties to commercial premises (see IHT, 1997).
4.37 The fact that a relatively high proportion of bicycle accidents occur on links strengthens the case for
properly planned and designed facilities for cyclists, particularly in urban areas. While there is no single
solution to providing a suitable infrastructure for cycling, the hierarchy of measures set out in
Cycle-friendly Infrastructure (IHT et al, 1996a) should be carefully considered before choosing the design
solution. This includes:
traffic reduction;
traffic calming;
junction treatment and traffic management;
redistribution of the carriageway; and
cycle lanes and tracks.
If road links remain with heavier vehicles or with a high speed differential between cyclists and other road
users, then the case for segregation on-carriageway or off-carriageway is strengthened (see Appendix A8).
4.38 For school children the safe routes to school approach is proving to be very successful (Paragraph
4.176 below and Appendix A18).
- 43 -
4.40 These roads have proved difficult to treat, partly due to the complexity and potential conflict of the
activities and partly due to the perceived need to maintain high capacity for traffic flow, and to maintain
relatively high levels of speed. In other words, the traffic function has been given priority over other
activities.
4.41 Mixed Priority Routes are generally A or B class urban roads, and frequently radial roads which may
pass through town centres. The speed limit is usually 30 mph, occasionally 40 mph. They are typically the
main roads into and out of towns, and it is rarely feasible to provide alternatives to them.
4.42 Owing to the lack of specific definition in STATS19 accident data, it is difficult to determine a
national figure for the number of accidents on Mixed Priority Routes, but 30 per cent of all casualties
occur on urban A roads. Vulnerable road user groups of pedestrians and two-wheeled vehicle riders,
prevalent in town centres, have casualty rates many times higher than those of car occupants (see IHT,
1997).
4.43 The benefits from reducing speeds on this type of road are particularly high offering up to a 7 per
cent reduction in accidents per mph reduction in average speed (Taylor et al, 2000).
4.44 The most successful approaches to treating Mixed Priority Routes have involved instigating a change
in road environment using a combination of measures to manage speed to appropriate levels, and to
allocate different parts of the road space for the different functions. In particular the needs of the
vulnerable road user have been a high priority. The through traffic function is still catered for but in a way
that is compatible with the other users.
4.45 Consider the following measures when treating a Mixed Priority Route:
separating the through-flow, distribution and access functions(Where there is insufficient width for
separating functions, the through-flow function must be downgraded in priority);
raising the priority given to pedestrians and cyclists, and giving them specific space such as cycle
lanes and wider foot ways (LTN 2/86;LTN 1/89);
using gateways to emphasise the transition from one type of road to another;
reducing the difficulties of certain manoeuvres and preventing unsafe manoeuvres; and
using narrow lanes and channelisation. (Care should be taken to ensure that provision for cyclists is
still a consideration).
4.46 Depending upon the available road and pavement width, there are three broad types of treatment that
can be applied to Mixed Priority Routes:
full separation of functions;
partial separation of functions; and
one way solution for narrow roads, though appropriate traffic calming measures may also be
essential to prevent an increase in speeds.
(See Appendix A2, A8, A10, A19 for examples).
4.47 Town centre roads also often include bus facilities. As with all other traffic management measures,
features designed to ensure bus priority have to be considered in the context of safe design and operation.
Clearly safety risks are minimised if full physical separation from other traffic can be achieved. However,
- 44 -
- 45 -
Figure 4.6b
- 46 -
4.56 26.2 per cent of all casualties occur at minor urban road junctions, which are likely to be largely on
residential roads. The problems are the complexity and uncertainty of vehicle movements, especially
turning traffic, too high a speed of the straight ahead traffic, masking of vehicles, and the interaction of
pedestrians and two wheelers. Overall numbers of vehicles and pedestrians are generally low and their
presence by an individual road user often unexpected. Accident numbers at any one junction are usually
low, and so consider the cost effectiveness of options carefully.
4.57 12.3 per cent of all casualties occur on urban minor (residential) road links. As at residential road
junctions, the proportion of children involved is higher than on major roads. The problem is usually one of
inappropriate speed, and the presence of more vulnerable and inexperienced road users child pedestrians
and child cyclists.
4.58 In recent years it has been recommended that efforts are made to eliminate through-traffic on
residential roads using Urban Safety Management techniques, as described in 4.8. Speeds of the
remaining traffic are then commonly reduced through the introduction of traffic calming measures, either
as free standing schemes or as 20 mph zones (TAL 9/99; Appendix A23).
4.59 In 20 mph zones, appropriate speeds can be achieved through a combination of:
road humps (Appendix A12);
speed tables (Appendix A12); and
horizontal deflections (such as chicanes, throttles or narrowing Appendix A6).
4.60 Traffic calming on links also reduces speeds on the approach to junctions. Extension of the foot way
across the mouth of a side road by using a flat-topped hump (sometimes referred to as a foot way
crossover), can help to slow turning traffic at a junction and deter through traffic (IHT, 1997; County
Surveyors Society et al, 1994a).
- 47 -
4.61 To further break up the speed of traffic, mini-roundabouts may be used at busier junctions (Appendix
A15).
4.62 The use of speed cushions (Appendix A21) is appropriate where speeds are not required to be so low
as in 20 mph zones, and/or on bus routes or through routes for emergency vehicles.
4.63 The speed achieved by traffic calming measures is closely related to the spacing of the measures.
4.64 Chicanes have been used less extensively than vertical measures and there is more variability in their
level of acceptance by the public(Webster, 1998).
4.65 20 mph speed limits without self-enforcing measures have only a minimal effect on vehicle speeds
(see Appendix A23; DTLR, 1999c).
4.66 Physical traffic calming has achieved good reductions in casualties 60 per cent reduction for all
casualties but 70 per cent reduction for children (Webster and Mackie, 1996), though a little of this effect
results from some re-distribution of traffic.
4.67 On residential access roads drivers need to be given visual cues that indicate strongly that this road
space is part of the environment where people live, walk, talk and play.
4.68 A way to treat such roads may be to create Home Zones (first installed successfully in Holland as
Woonerfs). In these zones, the full road space is very much shared between motorised and non-motorised
users. A pilot programme of nine schemes of different types is currently taking place in England and
Wales (Layfield, 2000).
4.69 The speed of traffic is kept very low by the intricate nature of the street layout, the placement of
street furniture and features, and by generating local ownership of schemes. However, these schemes are
generally very costly.
Rural roads
Figure 4.7a
- 48 -
Figure 4.7b
4.70 From Figs 4.7a and 4.7b it can be seen that the vulnerable road user groups on rural roads are again
prevalent. As you might expect, they comprise higher proportions of the fatal than all-casualty category
totals (though the proportions of pedestrian and pedal cycle casualties are lower than their urban
equivalent, probably reflecting lower levels of walking and cycling).
- 49 -
4.71 A number of recent publications address rural accident problems and solutions separately from urban
ones. For example, DTLR (2000a); IHT, 1999c; Barker (1997); Barker et al (1998); Barker et al (1999);
Gardner and Gray (1998); Hughes and Amis (1996); Hughes et al (1997);Pickering et al (1986); and for
villages, reports relating to the VISP study such as CSS and DOT (1994b); Wheeler and Taylor
(1995);Wheeler and Taylor (1999); Taylor and Wheeler (2000). In addition, note that many of the Design
Standards etc in the bibliography relating to trunk roads, may be useful in the design of local roads,
particularly major rural roads.
4.72 More than half of all fatalities and a third of all seriously injured casualties occur on rural roads
(speed limit 50 mph and above) see paragraphs 3.36 -3.40 above. Accident severity is higher on rural
roads than in built-up areas (speed limit 40 mph and below), and vulnerable road users (pedestrians and
cyclists) are particularly at risk. These results are, in part, likely to be due to generally higher vehicle
speeds on rural than on urban roads. But of crucial importance is the fact that rural accidents are generally
more thinly spread over a wider area than accidents in towns. Cost-effective treatment to prevent
accidents is therefore more difficult to apply and the best locations for treatment harder to identify.
4.73 Figures 4.8a and 4.8b break down all casualties and all fatalities on rural and urban roads by casualty
type. Figure 4.8b shows that almost a quarter of all pedestrian fatalities and almost a half of all pedal
cyclist fatalities occurred on rural roads. These figures are worrying given that walking and cycling are
mainly associated with urban travel.
Figure 4.8a
Figure 4.8b
- 50 -
4.74 Two-thirds of accidents on rural roads occur on single carriageway roads. The accident rate (per
vehicle-km) is higher on single carriageway roads than on dual carriageways and motorways.
4.75 TRL Report 304 (Barker et al, 1998) describes an analysis of the characteristics of injury accidents
that occurred on all rural single carriageway roads in Great Britain in 1994-95, based on the
nationalSTATS19 data-base.
4.76 Rural safety management should involve:
identifying a functional hierarchy of roads and encouraging traffic on toll roads with an appropriate
function; and
managing vehicle speeds at the right level for the conditions on each type of road in the hierarchy.
4.77 Re-distributing traffic onto appropriate roads. Where alternative routes exist which enable heavy,
through traffic to avoid villages, for example, or where rat-running is a problem, measures should aim to
slow the traffic and thereby discourage the use of inappropriate roads. Other modifications may be
necessary on alternative routes to cater for the additional traffic.
4.78 Managing vehicle speeds. On rural roads, speeds that are too high for the conditions are likely to be
more of a problem than speeds in excess of the speed limit. High accident severity rates will be improved
by reducing speeds. Vulnerable road users will particularly benefit.
4.79 Managing vehicle speeds will involve setting appropriate speed limits and the use of engineering and
enforcement measures. Appropriate speed limits will be determined by many factors, including:
- 51 -
Rural Villages
4.85 Local councils and residents have voiced concern over many years about traffic nuisance and
perceived safety problems in villages. This is normally expressed in terms of too much traffic travelling
too fast through the village.
4.86 The scale of the national problem is not easy to define since it is difficult to extract the relevant
accident data from the national STATS19database for all accidents in villages throughout the UK. Many
villages have more of a perceived problem than a real one, as the numbers of injury accidents are often
small. Nevertheless, there is often a case for the implementation of traffic calming type measures, as
- 52 -
reductions in traffic speed can generally be expected to lead to reductions in accidents and casualties
cost-effectively (Taylor et al, 2000; TAL 11/00).
4.87 Where traffic calming measures have been introduced in villages, they have reduced mean speeds by
up to about 10 mph. Accidents (particularly the most severe accidents) have substantially reduced overall
(Wheeler and Taylor, 2000; Taylor and Wheeler, 2000).
4.88 The differential between the speed limits inside and outside a village can be large. If drivers have
been travelling along rural roads subject to the national speed limit for an appreciable distance, they may
not recognise the need for greater care and lower speeds. They may be unaware of a lower speed limit or
of their own speed and may respond late to the lower limit. In particular they may be unaware of the
increased risk of an accident, especially with a vulnerable road user. Speeds observed through villages are
often high compared to what is appropriate for the conditions.
4.89 The increased number of pedestrian movements and the greater concentration of cyclist and
equestrian journeys within the environs of a village warrants special attention, particularly if accidents are
taking place during the hours of darkness.
4.90 The principal aim is to alert drivers to the change in road environment. Although village name signs
together with speed limit signs have been conventionally used to mark the entry to a village, a gateway
can make this change more prominent (see Appendix A9).Ideally the gateway and a speed limit change
should coincide with the boundary of the village, to provide all the visual cues together.
4.91 Gateways are generally not enough to produce speed reductions that are sustained throughout the
village. Other features within the village are required and these may include:
narrowings (eg pinch-points, build-outs see Appendix A6);
traffic islands and pedestrian refuges (see Appendix A11; TAL07/95);
coloured surfaces and markings (see Appendix A7 and A20; TSM Chapter 5, CSS, 2000; DMRB,
TA 81/99);
over-run areas at junctions (see TAL 12/93);
mini-roundabouts (see Appendix A15);
signs (see Appendix A20; TSM Chapters 3, 4, 5; and 7);
speed cameras, if cost-effective (see Appendix A4).
4.92 Consider the need to light potential hazardous features where village lighting is poor or non-existent.
4.93 Features need to be designed sensitively to minimise impact on the rural character of villages and be
located to minimise problems for local residents.
4.94 20 mph zones are only appropriate if supported by physical features that will ensure reduced speeds
are achieved and are therefore unlikely to be used on major through routes.
4.95 A history of pedestrian injuries will need close examination to establish common factors. If
pedestrians have been struck while walking along the road, check whether footways and perhaps lighting
are adequate. If they have been hit while crossing the road, then it is likely that special provision is needed
in the form of refuges and/or a zebra/pelican/puffin/toucan/pegasus crossing (see Appendix A10;
LTN1/95; LTN 2/95). Where such measures involve a reduction in carriageway width, cyclist and
- 53 -
equestrian needs should be considered and special provision made where feasible (see Appendix A8).
Motorways
Figure 4.9
4.96 The Highways Agency is responsible for motorways 37 , so for completeness they are included here.
Many of the problems and solutions are applicable on other major roads, especially dual-carriageways. In
addition, some local authority schemes may include motorway/non-motorway intersections or
roads/footways that cross, but do not intersect, motorways.
4.97 Almost 5 per cent of all casualties occur on motorways. Of these, about 6 per cent are children under
16 years of age.
4.98 Motorway junctions are relatively widely spaced and consequently only17 per cent of motorway
casualties are injured in accidents at junctions. Motorways carry more traffic than other types of road but
have the lowest accident rates and casualty severity indices 38 of all roads (11 accidents per 100 million
vehicle-kms and 11 per cent casualty KSI index, compared with 50 per 100 million vehicle-kms and13 per
cent casualty KSI index for all roads, respectively DTLR, 2000a). These statistics reflect the high quality
(dual-carriageway) design and build of motorway roads, the use of hard shoulders, the low junction
densities, one-way traffic flow and the low opportunities for pedestrian and pedal cycle conflicts.
4.99 However, when motorway accidents do occur, they frequently involve more than two vehicles and
result in a number of injuries. This may be a consequence of high quantities of traffic, high speeds and
vehicles driving too close together. On average, there are 2.27 vehicles and1.63 casualties per motorway
accident, compared with averages of1.83 and 1.36, respectively, for all accidents. It is partly for this
reason, and partly because of the distances involved for recovery and emergency vehicles, that motorway
accidents are the most costly.
- 54 -
4.100 Accidents on motorways are also more likely to involve only one vehicle (22 per cent compared
with an average of 14 per cent on all roads),perhaps as long periods of driving without a break and a lack
of visual stimulation can result in driver fatigue or distraction.
4.101 Motorways do not permit pedestrian or small two-wheeler vehicle traffic and so the largest
differentials between the speed and mass of individual vehicles will usually be between cars/TWMVs and
goods vehicles (HGV/Luvs) 39 . Most often it is the drivers and passengers in the cars/TWMVs that are
injured, however, when HGV/LGV occupants are injured, they tend to be severely injured (perhaps if they
are unrestrained by seat belts or are trapped in wreckage). About 38 percent of all motorway casualties are
injured as the result of an accident involving at least one HGV or LGV, and these casualties account for
around half of all killed or seriously injured motorway casualties.
4.102 Although the vast majority of motorway casualties are car occupants (about 85 per cent),
motorcyclists have accident rates about 10 times higher than the average for all road users (per vehicle-km
travelled).4.103 Almost all children injured in motorway accidents (94 per cent) are car occupants. Of
these, 78 per cent were rear seat passengers. Information regarding rear seat-belt wearing compliance by
these children is not known.
4.104 The numbers of pedestrians injured in motorway accidents is small but still of concern, as exposure
is so low. Pedestrians are at particular risk on hard shoulders, at roadworks and at motorway junctions.
4.105 Accident rates in the vicinity of roadworks are higher than on similar road sections without works.
50 mph speed limits for these sections are now common practice and are often enforced with speed
cameras. The majority of accidents involve shunts. Drivers often drive too fast for the conditions, too
close, too aggressively or without due care and attention. For advice see Health and Safety Executive,
2000.
4.106 Close-following behaviour can be addressed through the use of appropriately worded variable
message signs which require special authorisation. Most of the motorway network is monitored by
cameras and once an incident, or congestion, is spotted the effects can be minimised by informing drivers
of the problem ahead and advising/encouraging/enforcing action or diversions 40 . Such signs can
comprise dot matrix symbols or include a worded message.
4.107 When a problem persists at a particular location, the provision of static signs (TSM Chapter 4)
should be considered and, if the traffic flow levels are suitably light, the use of chevron spacer road
markings might be considered. (Appendix A5).
4.108 The safety problems relating to fatigue and distraction and those regarding pedestrians and
motorcycles may best be addressed through non-engineering techniques but should still be borne in mind
whenever engineering work is carried out.
4.109 Roadwork sections need to be carefully designed see TSM Chapter8. Attention needs to be given to
the use, the extent, and the enforcement of a lower speed limit.
- 55 -
Rural Dual-carriageways
4.110 Many dual-carriageways, particularly the most recently built, are built to a high standard and
designed to service the traffic they carry. Consequently, the accident problems on many dual-carriageway
roads are similar to those observed on motorways (Walmsley and Summersgill, 1998). However, those
built to older, lower standards may have more potential for improvement.
4.111 Although only about 4 per cent of all accidents occur on rural dual-carriageway A-roads, these
accidents account for about 6 per cent of all casualties. The reasons for this are likely to be similar to
those suggested for motorways in paragraph 4.99 above.
4.112 About half of casualties are injured in accidents on links and about half at junctions. The proportion
at junctions is quite high when compared with that for all rural roads (38 per cent). The severities of rural
dual-carriageway accidents are slightly higher than those on motorways with20 per cent of accidents
resulting in fatal or serious injury (Barker et al,1999).
4.113 Of the accidents at junctions, about 54 per cent are at roundabouts. Many of the problems at other
junctions are a consequence of drivers difficulty in judging the speed and distance of other traffic,
particularly at priority junctions.
4.114 The approaches to junctions should be adequately and clearly signed(see TSM Chapters 4 and 7 and
TSRGD). This is of particular benefit to drivers on this type of road as safe opportunities to stop and
consult a map or turn around may be few and far between. The provision of anti-skid surfacing on the
approach to the junction may also be worthwhile (see Appendix A1; DMRB, HD 28/94).
- 56 -
4.115 Junctions (and traffic at or on the approaches) should be conspicuous and drivers should have
adequate warning to slow down and be aware of the path they should take through the junction. Priorities
should be clear. The provision of lighting may be appropriate, if a suitable power supply is nearby. Grade
separation of the junction maybe appropriate for the most major junctions.
4.116 The provision of yellow bar markings on the approaches to at-grade roundabouts is one option for
giving drivers advance warning of the junction ahead (only if the approach is high speed) and may be
particularly suitable on approaches with little visual stimuli. See Appendix A26.
4.117 On large roundabouts lane markings may help guide drivers and riders through the junction (TSM
Chapter 5). Two-wheeled vehicles often experience problems at roundabouts and their needs should
always be considered (Appendix A8 and A15).
4.118 It is necessary to ensure that all accesses and not just those with other main roads are safe and of a
suitable standard to accommodate the traffic using it. For example, laybys, private drives and businesses,
farm accesses, bus stops etc should be designed so that traffic can join and leave the main road safely.
4.119 The safety of vulnerable road users crossing junctions should always be considered as junction
widths can be very wide and traffic flow fast and heavy. The provision of signals with pedestrian phases
can provide opportunities for pedestrians and cyclists to cross one arm of a junction in several stages
(LTN 1/98; DMRB, TD 50/99). The provision of grade separated pedestrian, cycle and equestrian
crossings may also be appropriate if flows are high enough (see paragraph 4.127).
4.120 If junctions are uncontrolled and traffic needs to cross the two carriageways in two stages, care
should be taken that the central reservation holding area is sufficiently wide and that traffic on the main
road is clearly visible to drivers using the gap.
- 57 -
4.121 Where accident problems on links closely resemble those on motorways they should be treated
accordingly.
4.122 A consistent approach along a route (for example, to indicate the relative severity of bends) is
important (TSM Chapter 4; TSM Chapter 5;IHT, 1999).
4.123 In addition, care should be taken to ensure that hazards are adequately signed or marked in advance
so that drivers can adjust their speed or position accordingly.
4.124 In general, as many visual cues as possible of the changing alignment of the carriageway should be
provided at bends. These may include chevron boards, reflectorised posts, white lining etc (see TSM
Chapter4 and TSM Chapter 5). On major routes that are predominantly straight, even moderate bends may
need to be well signed (see TSM Chapter 4 and TSM Chapter 5).
4.125 The use of safety fences on the central reservation will substantially eliminate the opportunity for
head-on collisions. Safety fences at the roadside can be used to protect vehicles leaving the road on the
near side from hitting objects such as trees and lampposts and from going over embankments into gullies
by guiding them back onto the carriageway. Gaps in the central reservation should be kept to a minimum
and restricted to locations where they can be safely used.(See Appendix A13; DMRB, TD 19/85; DMRB,
TD 32/93.)
4.126 On routes with high volumes of HGV traffic the provision of climbing lanes uphill for overtaking
traffic and escape lanes downhill may be considered appropriate.
4.127 The provision of crossing facilities for non-motorised road users should be given consideration
(LTN 1/95; LTN 2/95). Although on busy roads the provision of subways or bridges may be justified for
non-motorised users, these must be carefully designed and positioned to encourage people to use them,
given past experience of their unpopularity. In some recent cases, road levels have been altered to allow
footways and cycle tracks to remain at ground level.
4.128 Depending on demand, consider the need to provide facilities for cyclists along rural dual
carriageway links. If there is a demand from pedestrians, then under the Highways Act, the highway
authority has to make provision for pedestrians. In certain circumstances it may be appropriate to provide
a shared use cycle track and footway segregated from the main carriageway. The principles set out in
LTN2/86 Shared Use by Cyclists and Pedestrians, will guide the designate the most appropriate solution.
- 58 -
4.132 Overall, these roads carry a similar amount of traffic per year as motorways do i.e. about 30 per
cent of all rural traffic (Barker et al,1999). However, at seven times the kilometre length of motorways,
the potential for major redesign is low on grounds of cost. Therefore, more cost-effective solutions have
to be adopted.
Figure 4.12
4.133 Older drivers are more likely to be involved in accidents at junctions than younger drivers.
4.134 The number of potential conflict points should be minimised; T-junctions have lower accident rates
than crossroads and multi-arm junctions should be avoided (IHT, 1999). The use of traffic signals should
also be avoided where possible. As speeds and flows increase, advance warning becomes essential.
4.135 Anti-skid surfacing on high speed approaches can be effective (see Appendix A1 examples; DMRB,
HD 28/94).
4.136 The presence of junctions must be made clear to drivers and riders with clear and consistent
advance warning signs and carriageway markings (TSM Chapters 4 and 5) and by the presence of
reflectorised posts, traffic islands and bollards. Through drivers must be alerted to the potential hazard of
emerging traffic and encouraged to slow down; this is especially true where traffic turning off the major
road impedes through traffic.
4.137 The design of turning facilities for major road traffic at T-junctions and crossroads is important.
Vehicles, particularly cyclists, are vulnerable when positioned between fast traffic in both directions.
Protected lanes for turning vehicles can reduce accidents, and conspicuous and consistent road markings
are essential see TSM Chapter 5.
- 59 -
4.138 57 per cent of rural major single-carriageway casualties are injured on links.
4.139 Accidents on links are more likely than those at junctions to involve a single vehicle, young drivers,
a pedestrian, bad weather, and (probably because vehicle impact speeds are, on average, higher on links)
to result in more severe injuries.
4.140 Following distances are particularly important where forward visibility is restricted by, for example,
hills and bends. The problem can be exacerbated by a wide carriageway which can encourage staggered
following behaviour with shorter following distances (IHT, 1999c).
4.141 The use of white lining about 1m or more from the edge of the road (to provide a hard strip) or
continuous centre hatching to reduce the effective carriageway width to, say, 7m is likely to reduce
accidents (TSM Chapter 5; IHT, 1999c).
4.142 Double white lines are used to indicate stretches of road with limited forward visibility (at hills or
bends) where it is unsafe to overtake. TSM Chapter 5 gives advice on usage.
4.143 Speed cameras are an increasingly realistic option to enforce speed limits when there is a speed
related accident problem (IHT, 1999c). However, enforcement in remote areas is a problem where the
availability of power to such sites is a factor (See Appendix A4).
4.144 Speed differentials of mixed traffic should be reduced as far as possible. Careful consideration
should be given to providing for the needs of pedestrians and cyclists and the principles set out in the
hierarchy of measures in Cycle-friendly Infrastructure (IHT et al, 1996a)will assist the designer to adopt
the most appropriate solution. Where space permits, segregation of these vulnerable road users from
motorised vehicles may be appropriate. In certain circumstances it may be appropriate to provide a shared
- 60 -
use cycle track and footway and the principles set out in LTN 2/86 should be followed. If the road is
frequently used by equestrians, the provision of bridleways, which can also be used by pedestrians and
cyclists, may be considered. Provision of climbing lanes up, and escape lanes with arrester beds down,
steep hills may be relevant where the traffic has a high HGV content and the required road width is
available.
4.145 Icy conditions and wet roads can be the catalyst for accidents involving skidding. Drainage and skid
resistance should be checked and, if conditions merit, remedial action and signing should be implemented
quickly when dangerous conditions arise.
4.146 For roads with steep drops close to the carriageway safety fences should be considered (DMRB, TD
19/85; DMRB, TD 32/93). Where there is livestock adjacent to the road, roadside fencing must be
well-maintained and secure to prevent animals from straying into the paths of vehicles.
42
4.147 Although only about 10 per cent of all casualties in GB occurred on minor rural single-carriageway
roads in 1999, these accounted for 15per cent of all fatalities.
4.148 The numbers of accidents per vehicle-km are also very high on these types of road. (45 accidents
per 100 million vehicle-kms, compared with 26 for all rural roads Barker et al, 1999).
4.149 This is to some extent likely to be due to inappropriately high speeds on such roads, especially on
links, leaving drivers little time to react and recover when emergency situations arise, frequently resulting
in very serious injuries.
4.150 In addition, today, these roads are often carrying large volumes of traffic, far in excess of the levels
for which they were designed. They are often built to older and lower design standards, and may not be
subject to the same level of maintenance programming as those of more major roads and motorways.
4.151 The increased traffic has created rural roads that are unsafe for cycling, walking or horse riding.
Efforts need to be made to reverse this perception especially on these minor roads if the aforementioned
more healthy, often leisure pursuits are to be encouraged in favour of dependence on the motor car (IHT,
1999c).
4.152 Higher proportions of accidents than expected occur at night-timetaking into account the relative
traffic levels during the hours of darkness compared with daylight (IHT, 1999c).
4.153 Rural locations are perhaps more likely to suffer from lower levels of regular maintenance than
more major roads. Road surfaces therefore may become more slippery in general and this can produce
problems for (especially two-wheeled) vehicle drivers and horse riders.
4.154 These roads are estimated to account for about 82 per cent of the total rural road length in Great
Britain (Barker et al, 1999). Therefore, it is especially vital to adopt low-cost accident remedial measures,
such as signing and marking in accordance with TSM Chapters 4 and 5.
- 61 -
4.155 In general, speed reducing measures and measures that act as alerting devices for hazards (eg
supplementary plates on signs or rumble areas (TSM Chapter 7; Appendix A17)), which may be obscured
or difficult to judge or detect, are recommended and many of these are discussed below.
4.156 The situation with respect to numbers of casualties at junctions and on links tends to be reversed on
minor roads in the rural environment in that fewer numbers occur at junctions, ie a third at junctions
compared with two thirds on links. This may be largely due to the lower numbers of junctions and turning
vehicles than in urban areas. Nevertheless over100 people were killed in 1999 at rural minor road
junctions.
4.157 There can be a particular problem at crossroads where the road ahead, but not the junction, is
clearly visible to those approaching drivers who need to give way or stop at the junction.
4.158 Junctions where minor road vehicles must give way to major road traffic need to be visible. Where
skidding is evident or there are large numbers of recorded rear-end collisions, then there may be a case for
special anti-skid surfacing to be laid in the approach to a rural junction (Appendix A1; DMRB, HD
28/94).
4.159 Vehicle-activated signs which warn drivers who are approaching too fast, of the junction ahead may
be applicable at junctions with crossover accidents and high vehicle speeds on the approaches, if existing
signing is of a high standard and if a power source is available (Appendix A25).
- 62 -
4.160 Although only about 7 per cent of all casualties injured on roads in GB in 1999 were injured on
minor rural single-carriageway roads, these accounted for about 12 per cent of all fatalities. This gives
some indication that the severity of accidents on rural minor roads tends to be generally higher than on
other roads, which may be largely attributable to vehicle speeds often being generally too high on such
roads.
4.161 Many accidents on rural minor roads tend to be associated with vehicles leaving the carriageway
following a loss of control on bends.
4.162 Physical features (vertical/horizontal deflections such as road humps/narrowings) will be largely
inappropriate on, even minor, rural roads that carry high speed traffic. One aim of signing/marking is to
convey the impression of a hazardous situation for example, markings can be used to give the impression
of the road narrowing such as channelisation or edgelining in accordance with TSM Chapter5. These are
perceptual techniques.
4.163 On minor unlit rural single-carriageway links without kerbs, TSM Chapter 5 recommends white
lining systems. Such systems will include edge-lining and centre-lining using solid, dashed, or double
lines as appropriate and as specified in TSM Chapter 5 see Appendix A20.
4.164 On wide roads, measures that appear to reduce the road width (eg centre channelisation) may be
suitable to discourage overtaking and encourage lower speeds. The introduction of cycle tracks could also
be considered to provide some degree of segregation between cycles and other traffic.
4.165 Options for making provision for equestrians, include bridleways, all purpose highways without
motor vehicles 43 , or a margin at the side of the highway 44 . These may also help pedestrians and
cyclists.
- 63 -
4.166 As mentioned above, a predominant type of accident site is bends. Therefore, it is recommended
that measures suitable for use at or on the approach to bends should be considered, perhaps as a mass
action approach if the problem is widespread and existing signing and markings comply with advice in
TSM Chapters 4 and 5.
4.167 Vehicle-activated warning signs may be appropriate on the approaches to particularly hazardous
bends. They alert drivers and slow and smooth vehicle speed profiles through a bend. If a power source is
not easily available, the cost may be prohibitive (Appendix A25).
Non-engineering measures
4.168 Although this guide is primarily concerned with engineering good practice, modifying the road
environment should not be done in isolation and the approach to improving road safety must be an
integrated one. There are other very important areas where road safety activity should be carried out in an
organised and efficient manner, often co-ordinated by road safety officers.
- 64 -
4.175 Some of the more recent strategies that are relevant to the latest national casualty reduction targets
are discussed below.
- 65 -
4.182 A school travel plan works by looking in detail at childrens needs on the school journey and can be
geared to the needs of a primary or a secondary school. Ideas often include a rota for parents to
accompany younger children on a walking bus, cycle stands at the school, cycle training, and low fare
deals for children using public transport. A school safety zone can transform childrens journeys with
crossing points, traffic calming and lower speed limits on nearby roads.
- 66 -
Cycle training
4.194 Road sense training for child cyclists has been provided for many years, normally out of school
hours, and is regarded as a valuable exercise.
4.195 Savill and Bryan-Brown (1996) evaluated eight schemes around the UK. Groups of 13-year old
children who had received training at age11 were given a cycle riding and knowledge test by the local
road safety officers. The study concluded that these children did indeed possess significantly safer riding
skills and knowledge than those who had not taken part in the cycle courses.4.196 A relatively new
extension to cycle training carried out by some authorities (such as Surrey County Council) is safe cycling
classes held for both parents and children. This also provides adults with useful points to be wary of when
cycling with young children.
Pre-driver training
4.197 There is recent concern for the mid-teenager group. For example, the number of casualties killed or
injured in cars appears to begin to rise sharply from the age of 14, with more than twice the numbers of
casualties recorded for 15-year olds than any other single year child age-group under 14. Also, the 12 to
15 year age band is the only one where girls are proportionately more involved than boys.
- 67 -
Figure 4.16
4.198 Young drivers in the 17 and over age groups, continue to be a major problem, with casualty
numbers of 18-year olds being about eight times higher than most of the single year child age-groups.
4.199 There appears to be a need to focus on both trying to instil responsible attitudes to driving while still
in school. It is also advisable to get new immediate messages across to girls. For example, they should be
told that they need to be aware when travelling with their young friends when these drivers are taking
unacceptable risks, and ways in which they can try to modify such behaviour.
Publicity
4.202 Publicity campaigns are expensive but can be extremely effective, even though they may take many
years to bring about a lasting change in attitude (such as the well-known change in public attitude towards
drinking and driving). A study of evaluated campaigns has concluded that a well-designed publicity
campaign can typically produce a 30 percent reduction in casualties in the target group (Delhomme et al,
- 68 -
1999).
4.203 Recent areas where it has been identified that there is a need for publicity campaigns to be focused
(DTLR, 2000b) are:
improving child road safety;
preventing speeding;
drink and drugs driving;
drowsiness;
motorcycles;
use of cycle helmets;
mobile phones; and
company car drivers (for both employers as well as employee drivers).
32
Urban (or built-up) roads are defined as those with speed limits of 40mph or less.
33
Rural (or non-built-up) roads are defined as those with speed limits of 50mph or more. Roads through
villages with speed limits of 20, 30 or 40mph are not included. A rural road may or may not have
buildings alongside it.
34
The police operate speed cameras. Local authorities must liaise with the police where the use of
cameras is proposed.
35
Over recent years, some authorities in London have introduced roadside or bus-mounted cameras to
identify offenders.
36
In STATS19 data these have been assumed to be non A class single-carriageway roads in built-up
areas with speed limits of 40 mph or less.
37
Some local authorities are responsible for motorway standard roads and some are Highways Agency
Agents.
38
Casualty severity index (Casualty KSI index) = 100*Number of killed and seriously injured
casualties/All casualties.
39
TWMV = Two-wheeled motor vehicle; HGV = Heavy Goods vehicle; LGV = Light Goods Vehicle.
40
On the M25, an interactive system of variable speed limits has been introduced. Speed limits change
according to the prevalent traffic conditions ahead.
41
42
These have been defined as all non-motorway or A-roads with speed limits of 50mph or more.
43
- 69 -
44
5. Measuring effectiveness
Monitoring matters
5.1 Through the LTP Annual Progress Reports authorities will monitor the progress that they are making
in working towards the achievement of their local road safety targets contained in their Local Transport
Plans(see Chapter 3). In addition, authorities are expected to monitor the effects of individual safety
schemes. The following Chapter sets out advice on measuring the effects of such schemes.
5.2 What is required is some knowledge of how driver behaviour changes following the introduction of a
scheme and, ultimately what are the effects of behavioural changes on accident frequencies and casualties.
5.3 Monitoring these changes is the only valid and objective way to be able to demonstrate the relative
(cost-) benefits, and success in saving casualties, between more than one safety scheme. The results of
monitoring feed into future work, rejecting less successful types of schemes in favour of more successful
ones, or helping to make decisions about a number of small schemes over a single costly one. Ultimately,
this should produce greater accident reductions and fewer casualties.
5.4 The County Surveyors Society and Highways Agencys MOLASSES database stores information on
safety schemes installed on local and trunk roads across Great Britain. The database can be interrogated to
obtain information regarding the accident reductions achieved across all sites of a certain type eg
roundabouts, road humps etc. See the web site at www.trl.co.uk/molasses and Appendix C for an example
data entry form. However, note that TRL Limited, who manage the database, will accept data in almost
any format.
When to monitor
5.5 The monitoring studies for measuring the effect of a safety scheme are usually by before and after
analysis of factors that are likely to have a bearing on the safety of road users at the particular treated
site(s).Although not an exhaustive list these may include:
spot speed;
speed variance;
traffic conflict studies;
traffic volumes;
journey time/delay;
compliance with traffic control devices;
skid resistance;
sight line/passing sight distance/super elevation;
pedestrian safety gaps/kerb delay/crossing times;
road accidents; and
weather/season.
- 70 -
5.6 Before measurements should be made as close as possible to the time when the scheme is
implemented. Ideally, this would be during the month before (except for accident monitoring see
paragraph5.33). However, in some cases it may be sensible to make observations only after the changes,
for example, with attitude surveys.
5.7 It would, of course, be impractical to carry out detailed behavioural studies for all minor alterations,
but studies may be particularly important for expensive schemes like area-wide or mass action treatments.
It must be noted, however, that the behavioural or geometric variables listed above have the disadvantage
that they do not give a direct measure of the magnitude of safety improvement since the precise relation to
accidents is uncertain. However, despite this drawback objective measurements are often considered very
worthwhile, since they can give a good indication of a change in safety.
5.8 Measurements should not be taken during the installation period. Additionally, after installation, a
week or more should be allowed as an adjustment period for road users to become familiar with the new
scheme.
5.9 After measurements should commence within one month of site work being completed. It is often
desirable to take several sets of after measurements, at various time intervals after the scheme is
introduced, to investigate the extent to which any initial effect is sustained and to allow for seasonal
variations.
5.10 Where possible, monitoring should take place under normal traffic conditions and not coincide with,
for example, school and bank holidays, market days, early-closing days, poor weather or roadworks.
5.11 When comparing data between sites, it is useful if all monitoring can be carried out at the same time,
or if careful cyclic monitoring techniques are planned in the design of trials. If several schemes with
different installation periods are involved then consideration could be given to planning equivalent
monitoring periods with respect to installation dates.
5.12 It should be remembered that before monitoring can never be repeated! It is important, therefore, to
check before data before a scheme is installed.
5.13 It would be disappointing, to say the least, if there was not an immediate and noticeable improvement
in driving behaviour at a scheme (eg particularly a reduction in speed in, say, a traffic calming scheme).
What is more important, however, is that a worthwhile underlying improvement (that results in casualty
savings) remains after any initial novelty effect has worn off. It is this underlying improvement which is
the most important to measure. Experience from earlier research suggests that changes in behaviour
should have stabilised by12 months after installation and this is, therefore, recommended as a suitable
period to judge the value of the scheme in behavioural terms.
5.14 The ultimate measurement to consider is the effect of the scheme on accidents and casualties. The
main (but probably not the only)justification for introducing a scheme will probably be to improve safety.
Many schemes are designed to achieve reductions in vehicle speed and, given the now well-proven
correlation between speed and accident reductions (Taylor et al, 2000), one might reasonably expect
accidents to be reduced also. However, when monitoring only one scheme or a small number of schemes
over a short time, accident monitoring alone will only be a weak indicator, as it is most unlikely that small
numbers of accidents and short time-scale of the monitoring will allow any changes to be statistically
significant.
- 71 -
5.15 The extent of before and after measurements required is considered separately for each of the
monitoring options discussed below.
Control sites
5.16 In a perfect monitoring trial, there would be an equivalent site to the treated site (the Test site) at
which no changes were made throughout the monitoring period. Data from this perfect Control site would
be entirely representative of what might have happened at the Test site had the scheme not been
introduced. It would allow a more accurate comparison between before and after data, taking account of
any general changes in driving behaviour, travel patterns, weather, economic activity, etc.
5.17 Unfortunately, it is rare indeed to find an individual site that matches the Test site in all respects, but
Control sites should generally have similar features, traffic levels and traffic mix to Test sites. Generally,
attempts by researchers (eg Hauer, 1992) to find suitable Control data for trial sites has led to difficulties
in the interpretation of unstable data. In addition, collection of detailed Control data may add significantly
to the cost of monitoring. If the Test site is a route or area, then similar route/s or area/s may be suitable as
Controls. If the Test site is small, then using a large Control may be worthwhile because, by combining
data from many sites, any fluctuations at individual sites will tend to cancel each other out giving a more
accurate overall picture. This may include using data for a whole town, county, or even readily available
national data covering the same period; eg if a junction is improved, then the control group may be all
(similar type) junctions within the county.
5.18 In some cases, it may not be necessary to collect Control data, but to make the assumption that
conditions at the Test site would have changed little, if at all, during the period of the trial. Thus the
Control would be no change, and the after data directly compared with the before data.
5.19 It is most desirable to have a Control in situations where outside influences are thought to have
affected the Test site, for example, when a major external change occurs at the same time as the
implementation of the scheme, or during the monitoring period. An example of such a change might be
the opening of a parallel route that diverts traffic from the route through the scheme. If the scheme is
installed, then an appropriate Control must be found, or direct before-after comparisons will be
misleading.
5.20 In some instances, if a newly installed measure targets traffic in only one direction, traffic travelling
in the other direction may provide suitable Control data. However, it should be noted that drivers might
drive through the site in both directions and be familiar with the measure, which may affect their
behaviour.
5.21 Alternatively, in some instances, it may be suitable to use data from vehicles upstream of a measure
as control data. However, again, drivers may be familiar with the measure if they have travelled through
the site on a previous occasion, and this may affect their behaviour.
- 72 -
each scheme, which of the engineering measures in the scheme will be monitored separately, in addition
to the assessment of the effectiveness of the whole scheme.
Data quality
5.23 The conclusions of any monitoring programme are only as good as the reliability of the data on
which they are based. Experience has shown considerable variation in the quality of data collection even
when collected against a prior written specification. As a consequence, for data where an element of
choice or selection exists on the part of the data-collector, a site visit by the agent responsible for
commissioning the monitoring is always recommended to be able to brief the data-collector carefully on
his or her duties.
5.24 Examples where data quality can be impaired through inadequate briefing include:
collection in inappropriate weather conditions;
collection at inappropriate locations;
measuring both directions of travel (without identifying each measurement), when only one is
required or both required separately; and
not collecting sufficient, or even any, before data.
5.25 Where relevant, the same equipment and, preferably, personnel should be used for before and after
monitoring to ensure the consistency of results.
5.26 Experience also suggests that automatic equipment should be checked more than once a week to
ensure it is continuing to operate correctly and has not been vandalised.
5.27 In the case of attitude surveys, it is desirable that the commissioning agent attends the interviewer
briefing meeting to maintain consistency of approach and hence quality of the data collected.
5.28 Data that can be collected automatically must still be analysed consistently. Careful specification,
briefing and supervision of the analysis will be essential to obtain reliable results.
5.29 Back-up plans should be in place in case things do go wrong. It is recommended that a contingency
element be included in the monitoring budget in case of such problems (eg vandalism, theft, bad weather).
- 73 -
5.32 Early results are often very encouraging, but most innovative measures suffer from a novelty effect,
which reduces with time. It is therefore important not to rush to the residents, press, councillors or
pressure groups with these early results, as they are unlikely to be sustained. Ideally, wait for the 12
months after measurements to be analysed before announcing results.
Accident monitoring
5.33 For every safety scheme installed, the change in injury accident rates between the before and after
periods will be a major consideration. Normally, all injury accidents which occurred in a period of three
years before and three years after the introduction of the measures would be considered, but a preliminary
look at the data one year after would be valuable.
5.34 Accident data should always be collected and examined for changes in the accident categories that a
scheme is addressing. However, it is highly unlikely that a statistically significant result would be
produced at a single site. This is much more likely to be achieved if the data from several schemes are
combined and there are large changes.
5.35 The monitoring database MOLASSES addresses this problem by gathering together accident and
design information from all local authorities and Highways Agency agents. By pooling all the information
available for a particular type of scheme, a more robust picture of effectiveness can be achieved (see
paragraph 5.4 above and Appendix C).
5.36 Many of the monitoring options discussed below are proxies for accident monitoring.
5.37 Monitoring accidents is important to ensure that the scheme has not introduced a new problem. It is
also important as a reference tool to allow more accurate assessments of future benefits, when similar
measures are used at other locations.
5.38 Statistical tests for analysing accident data are outlined from paragraph 5.113 onwards.
- 74 -
5.40 The first two options involve spot measurements of speeds. These can give valuable information
about changes in behaviour at one point, or about elements of the scheme, but not of the scheme as a
whole. The third option, average speeds calculated from journey time measurements (over a measured
distance commonly referred to as journey speeds), is an excellent method of assessing the effect of the
scheme as a whole, particularly for a large scheme, a village scheme, or a route scheme.
5.41 If the speeds of cars and heavy vehicles are to be measured separately, then radar or laser
measurements may be the most appropriate method as automatic equipment generally only gives a crude
breakdown of the mix of traffic in the overall sample. A combination of all three forms of measurement
would provide the most complete picture but is unlikely to be financially viable.
5.42 Monitoring equipment should be made safe and secure. This will often involve chaining data loggers
and cameras to lampposts or installing unobtrusive lockable cabinets at the roadside.
5.43 As mentioned in paragraph 5.13 above, with any new road engineering measure, an initial effect on
speeds is to be expected. Frequently, the initial effect is greater than the longer-term (underlying) effect,
observed once drivers have become used to the new measures. One would therefore expect a gradual
return towards the before level of behaviour, after the introduction of any novel scheme. It is important to
be sure that the underlying effect of the measures is a lasting improvement and speed data should ideally
be collected at intervals over a period of twelve months to assess any relaxation towards before levels.
5.44 The most commonly studied characteristics of speed used to identify changes are mean speeds and
85th percentile speeds 48 . Research (Taylor et al, 2000) has shown that the most important determinants
of accidents are:
the mean speed;
the variability of speed; and
the percentage of vehicles exceeding the speed limit, and the margin by which they do so.
Accident predictive relationships are available which use either the first two or the second two of these
measures. Equipment that records individual vehicle speeds best allows these speed distribution
parameters to be determined accurately.
- 75 -
- 76 -
- 77 -
5.59 When results from more than one scheme are to be combined, some effort should be made to make
monitoring points comparable between schemes ie at similar features to allow cross-comparisons to be
made.
Flow monitoring
5.65 If there is an alternative parallel route to which drivers could divert in order to avoid a scheme, or
part of it, then flow is an important parameter to measure accurately. It is unlikely that drivers would
change their basic route for just one small scheme, but they might divert to parallel side roads if they saw
an advantage in this (eg to avoid driving over a series of humps). It is recommended that flows should
then be recorded and classified on both the original and alternative routes.
5.66 Background information on traffic flow data is likely to be available already for some roads,
particularly major ones. If this is not fully classified (ie detailing volumes of heavy vehicles, two-wheeled
motor vehicles and pedal cyclists separately), or up-to-date, then a classified flow count would be required
as part of the monitoring programme. However, flow data is often easy to collect as a by-product of, or in
conjunction with, other measurements (eg automatic speed monitors record flows as well as speeds).
5.67 As an alternative to automatic flow counts, observers at the roadside or on junction arms can carry
out manual classified counts (particularly for junction turning counts). The count period should not
normally be less than one hour but will depend upon flow levels and hourly, daily and seasonal variability
- 78 -
of traffic. A 12-hour period is often used. Scaling factors based on national averages may be used to scale
one hour flows (Highways Agency, 1996).
50
or inter-vehicle gaps
5.72 Monitoring headways is likely to be most useful in a context where increasing the gaps between
vehicles would be expected to improve safety (motorway chevron installations, for example). This needs
automatic vehicle logging using loop detectors.
5.73 The extent to which changes in inter-vehicle gaps makes it easier for pedestrians to cross the road is
not easy to assess directly. It is easier to examine this matter through pedestrian studies or attitude
surveys.
- 79 -
brakelights (or swerving). The observer assigns a defined severity grade to each conflict incident (see
Transport and Road Research Laboratory, 1987for details).
5.76 The process is relatively labour intensive, requiring teams of trained observers, and may only be
practical at junction-type locations owing to the limited distance over which observers can be expected to
monitor. However, at certain locations they can be a valuable diagnostic tool in helping to highlight the
circumstances and frequency in which road users are experiencing safety problems.
5.77 Conflict studies can also provide a means of evaluating a location before and after introduction of a
safety scheme. The data is usually expressed in the form of daily rates of particular types of conflict. The
rates are normally simply compared to answer the questions:
i) Has the remedial measure(s) successfully alleviated the problem(s)identified in the before study?
ii) Has the remedial measure introduced any undesirable secondary effects that may cause other safety
problems?
5.78 If a full statistically valid result is required then, as we are dealing herewith separate behavioural
events (though remember that there is always a degree of subjectivity in how they are recorded), the
frequency of conflicts can be analysed using the same methods as for accidents, as described below from
paragraph 5.110 onwards.
Attitude surveys
5.79 The views of drivers, other road users, residents and traders about schemes are important. Ideally,
some form of survey should be undertaken, preferably during the month after the installation of the
scheme, to provide general feedback. However, the objectives, size and type of the proposed survey need
to be carefully balanced against the costs of carrying it out.
5.80 Separate questionnaires for each of the target groups may be desirable.
5.81 The design of the questionnaire is very important. Questions must be clear and concise with open or
specified-choice answers required. They should address all the key points pertaining to the particular
scheme under evaluation. The questionnaire should be tested and designed for average completion in a
maximum of, say, five minutes for a street interview, and 20 minutes for a home interview or postal
questionnaire. Consideration may also need to be given to providing incentives for respondents to
maximise response rates and accuracy.
5.82 Such surveys may require prior approval (from elected members, for example). Sufficient lead-time
must be allowed for their consultation.
5.83 In most situations, only after attitude surveys will be needed in a monitoring programme, because it
is difficult to ask about a scheme which does not yet exist, except in general terms about existing
conditions. It is also likely that there might be a questionnaire survey to canvass local views as part of the
consultation process about the scheme itself. It would not be desirable to conflict with any consultation
process.
- 80 -
5.84 For most schemes it will also be appropriate to canvass the views of the emergency services, bus
operators and others after scheme installation (as well as at the planning stage as part of the consultation
process).
5.85 Remember that those who feel strongly are the most likely to respond to a postal survey, consultation
etc., and their views may not be representative of the majority.
Environmental monitoring
5.92 If it is expected that the measures proposed might have an effect on noise generated (when
introducing certain speed-reducing or alerting devices, for example), then some noise measurements
would be desirable. However, these are expensive and could only be justified with larger schemes. Overall
noise levels generally decrease with lower speeds, but the character of the noise may be affected by
vehicles crossing a measure (eg a road hump) and by greater acceleration or deceleration. Whether the
noise is a nuisance may also need to be assessed through a survey and complaints monitoring.
5.93 Both general background traffic noise and individual vehicle noise should be measured, both before
and after introducing the scheme. One set of measurements in each period should be sufficient to provide
a reasonable assessment of the change in noise.
5.94 Schemes that encourage a change in driving style may cause an increase in vehicle emissions.
However, if the scheme also encourages a decrease in traffic flow, this can counteract these increases. The
evidence so far suggests that the effect on air quality is likely to be small.
- 81 -
5.95 Quantifying the effect of a scheme on air quality is problematic because of the variability of
measured concentrations. There is an underlying downward trend in traffic emissions as new vehicles,
which comply with stricter emission limits, enter the fleet. Weather can also affect air quality, both
seasonally and year to year.
5.96 It is therefore important that surveys are designed carefully. Before and after surveys should be
carried out for at least three months and at the same time of year. Monitoring sites should be chosen to
include those roads on which the main measures will be installed and the roads that you might expect
traffic to use in order to avoid them. The sites should be located close to the emission source (ie close to
the kerb) so as to be able to detect, with some confidence, the changes in air quality resulting from
changes in emissions as a result of changes in driving pattern or traffic flow. A control site, outside the
scheme and preferably away from main roads, should be included within the surveys in order to
distinguish between the changes in air quality brought about by the measures and those resulting from
cleaner traffic and differences in weather conditions occurring between sampling periods.
5.97 Local authorities have a duty under Part IV of the Environment Act1995 to review and assess air
quality in their areas. The UK Air Quality Strategy identifies eight pollutants that should be included in
local air quality management. Of these, nitrogen dioxide (NO 2 ) and particles(PM 10 ) are probably of
most interest, as in some areas concentrations of these pollutants regularly exceed the current air quality
standards. Road traffic makes a large contribution to emissions of these pollutants, and so their inclusion
in routine surveys should be considered. Benzene and Carbon Monoxide should also be considered
because of their importance in terms of local air quality and also because they are largely derived from
vehicle exhausts. There are several methods of determining concentrations of these pollutants (for
example, diffusion tubes, automatic and battery operated samplers).
5.98 If pollution levels exceed set levels then an Air Quality Management Area has to be declared. In
these cases, local authorities are required to develop and implement air quality action plans, and it will be
all the more important to evaluate the air quality and other environmental effects of any new road safety
measures to minimise any conflict between them. It is necessary to adopt a balanced approach so that as
far as possible, measures which reduce accidents do not seriously prejudice air quality.
5.99 If a scheme causes a change in the number of parking spaces available, then it might be desirable to
survey parking habits on the main and surrounding roads. Before and after data, would be required and,
therefore, an assessment of the likely effect of the scheme on parking would have to be made at the
planning stage. Two monitoring options might be to examine the number and the percentage of spaces in
use over the period. Less objective information could also be made available from a resident or road user
survey, in particular by canvassing any effect on traders.
5.100 Schemes planned for major roads designed to accommodate high flows may change the level of
severance between the land use on both sides of the road. Severance will be a function of the demand to
cross the road and the opportunities for being able to do so. This can be investigated through video
analysis of pedestrian behaviour and crossing patterns, but this approach is expensive to analyse and
changes in behaviour are likely to be small and difficult to identify, or prove. Severance is usually,
therefore, most appropriately investigated subjectively through the attitude survey of road users, residents
and traders.
- 82 -
5.101 The aesthetics of a scheme are extremely subjective. Also a particular measure in one location may
be generally welcomed, whereas the same measure might cause disquiet in a more
environmentally-sensitive area. It is important to consider aesthetics as part of the planning process and in
public consultations. There may be a conflict of interest, as drivers need clear, bold measures, which catch
their attention, while most residents are likely to prefer features that blend harmoniously with the local
environment. The easiest way to monitor perceived effects is through opinion surveys.
5.102 It is often extremely valuable to have drivers eye-view video and still photographic records of a
scheme both before and after the safety engineering measures are introduced. They are useful in resolving
subsequent queries about the changes and provide a ready means for interested parties to experience the
scheme without the need to visit it. They are also useful to provide illustrations for reports and
presentations, particularly of unusual or novel elements of a scheme.
Evaluation
5.103 Having devoted considerable effort and expenditure to improving hazardous sites, there is a need to
evaluate these improvements.
5.104 This section briefly outlines the evaluation of schemes based on some of the types of monitoring
already described. It mentions the simple statistical tests needed to interpret the results. Appendix B gives
further detail.
5.105 To obtain statistically reliable results for accident changes, it is normally necessary to wait several
years after introducing the counter measure or package of measures has been introduced. It is, however,
assumed that the user of this Guide will need to interpret accident and other data practically without
necessarily having a full understanding of the underlying statistical theory, which can be quite complex. It
is, however, important that the user is sufficiently confident with his or her analysis. If in doubt, it is
strongly advised that help is sought from a professional statistician.
- 83 -
- 84 -
how the accident numbers at a site change relative to control data(see Appendix B.3).
- 85 -
one year
7 to 15 per cent
two years
5 to 11 per cent
three years
5.123 It is suggested, therefore, that where the highest accident sites are chosen for treatment, then the
above levels of allowance should be made when quoting the actual reduction in accidents that the schemes
have produced. The way in which a more accurate estimation can be obtained is rather complex requiring
use of data from similar sites to the treated sites but a method described by Abbess et al (1981) is outlined
in Appendix B.6.
Accident migration
5.124 The existence of accident migration is a fairly controversial issue but has been reported to be a real
effect (Mountain et al, 1992; Boyle et al, 1984; Persaud, 1987). It is simply that an increase in accidents
tends to be observed at sites adjoining a successfully treated site, giving an apparent transfer or migration
of accidents. It is unclear precisely why this effect occurs but is hypothesised that drivers are
compensating for the improved safety at treated sites by being less cautious elsewhere.
5.125 Obviously to detect such an occurrence, you need to compare the accident frequencies before and
after implementation of a scheme and those for the surrounding area with a suitable control group.
5.126 However, research and practical evidence (eg Brindle, 1986; Websterand Mackie, 1996) have
demonstrated that local area traffic restraint schemes do not create a significant increase in accidents on
surrounding roads. Mountain (1998) has more recently concluded tha ta more likely explanation for any
observed increase is a reverse regression-to-mean effect arising due to bias in the selection of the
neighbouring sites.
Behaviour adaptation
5.127 The effect of road users tending to alter their behaviour following introduction of a new safety
improvement, is now generally more accepted than the original, more controversial, philosophy of risk
compensation or risk homeostasis theory. The latter suggested that road users maintain a fixed level of
accepted risk, and so will take more risks when given greater accident protection, for example, by seat
belts or anti-lock brakes.
5.128 However, Trimpop and Wilde (1994) concluded that accidents are not necessarily the result of
risk-taking desire, but more of an inappropriate action based on faulty risk assessment. The challenge for
the road engineer is to introduce schemes that minimise the chances of road users making faulty
assessments; for example, in ensuring consistency in road users expectations for the level of road surface
friction, super elevation on bends, design of junctions and so forth.
- 86 -
5.129 Grayson (1996) concluded that evidence that adaptive or compensating processes are seriously
reducing the effectiveness of safety measures is slender and poses little threat to current road safety
practice.
Economic evaluation
5.130 For every scheme, the evaluation should include an indication of the benefits actually achieved in
relation to cost. Even if the scheme has been designed to tackle a very specific target group of accidents, it
is normal practice to include all accidents at the site in a full evaluation, in case the measure has had the
unforeseen effect of increasing other accident types.
5.131 The previous sections have already outlined how the best estimate of the size of the effect of a
scheme (or group of schemes) on accidents can be determined. If the site was one of the worst blackspots
in the area, then make some allowance for the regression-to-mean effect.
* A more accurate figure would be obtained by including only maintenance costs in this year and also
increased journey time costs if this is applicable.
- 87 -
5.136 The overall economic effectiveness of a scheme is indicated by the Net Present Value (NPV), which
is obtained by subtracting the Present Value of Costs (PVC, which must also be discounted if spread over
more than one year) from the Present Value of Benefits (PVB). This technique is described in more detail
with examples in COBA (DMRBvol. 13); RoSPA (1995b).
45
46
When data are binned, only the average values of speed for the vehicles passing within each time
period are stored.
47
Free-flowing vehicles are those where the driver has a clear choice of speed and is not influenced by a
vehicle ahead.
48
The speed at or below which 85 per cent of the vehicles in a speed measurement sample set were
travelling.
49
Automatic individual vehicle speed logging equipment will provide data that (with extra technical
effort) could be disaggregated according to headway information to isolate data from free-flowing
vehicles.
50
A distance headway is the gap between the front of one vehicle and the front of the next vehicle at a
moment in time. A time headway is the gap in time between the front of one vehicle passing a point and
the arrival of the front of the next vehicle at that point.
6. Bibliography
The Traffic Signs Regulations and General Directions (1994)
Statutory Instrument 1994 No. 1519 Road Traffic London: HMSO
The Traffic Signs Manual (various dates; DTLR, London: TSO)
Chapter 1. Introduction*
Chapter 2. Directional Informatory Signs on Motorways and All-Purpose Roads*
Chapter 3. Regulatory Signs (1986)
Chapter 4. Warning Signs (1986)
Chapter 5. Road Markings (1985)
Chapter 6. Illumination of Signs*
Chapter 7. The Design of Traffic Signs (1997)
Chapter 8. Traffic Safety Measures and Signs for Road Works and Temporary Situations (1991)
- 88 -
* To be published
[Note: Current advice relating to speed limit signs is given in TAL 1/95]
Traffic Advisory Leaflets
TAL 10/00: Road humps: discomfort, noise, and ground-borne vibration
TAL 11/00: Village Traffic Calming reducing accidents
TAL 12/00: Urban Street Activity in 20mph Zones Ayres Road Area, Old Trafford
TAL 06/00: Monitoring Walking
TAL 05/00: Traffic Calming Bibliography
TAL 04/00: Cycling Bibliography
TAL 03/00: Walking Bibliography
TAL 02/00: Framework for a Local Walking Strategy
TAL 01/00: Traffic Calming on Major Roads
TAL 17/99: Code of Practice for Traffic Control and Information Systems
TAL 16/99: The Use of Above Ground Vehicle Detectors
TAL 15/99: Cyclists at Road Works
TAL 14/99: Traffic Calming on Major Roads: A traffic calming scheme at Costessey, Norfolk
TAL 13/99: Historic Core Zone: Bury St Edmunds
TAL 12/99: Cycling for Better Health
TAL 11/99: Improved Cycle Parking at South West Trains Stations in Hampshire
TAL 10/99: Cycling Initiatives Register
TAL 09/99: 20 mph Speed Limits and Zones
TAL 08/99: Urban Safety Management: Using SafeNET
TAL 07/99: The SCOOT Urban Traffic Control System
TAL 06/99: Bikerail Combined Journeys by Cycle and Rail
- 89 -
- 90 -
- 91 -
- 92 -
- 93 -
HA 58/92 The Road Corridor [Amendment No. 1 Retaining Walls (Chapter 3)February 1997]
HA 59/92 Nature Conservation [Amendment No.1 Badgers (Chapter 5.3)February 1997]
HA 60/92 Heritage
HA 61/92 Contract and Maintenance Implementation
HA 62/92 Widening Options and Techniques
HA 63/92 Improvement Techniques
HA 65/94 Design Guide for Environmental Barriers
HA 66/95 Environmental Barriers Technical Requirements
HA 67/93 The Wildflower Handbook
HA 68/94 Design Methods for the Reinforcement of Highway Slopes by Reinforced Soil and Soil Nailing
Techniques
HA 70/94 Construction of Highway Earthworks
HA 71/95 The Effects of Highway Construction on Flood Plains[Incorporating Amendment No. 1
(August 1998)]
HA 72/94 Use and Limitations of Ground Penetrating Radar for Pavement Assessment
HA 73/95 Site Investigation for Highway Works on Contaminated Land
HA 74/95 Design and Construction of Lime Stabilised Capping
HA 75/95 Trunk Roads and Archaeological Mitigation
HA 78/96 Design of Outfalls for Surface Water Channels
HA 79/97 Edge of Pavement Details for Porous Asphalt Surface Courses
HA 80/99 Nature Conservation Management in Relation to Bats
HA 81/99 Nature Conservation in Relation to Otters
TA 8/80 Carriageway Markings. Markings for Right Turning Movements at Cross Road Junctions
Northern Ireland Addendum applicable for use in Northern Ireland
TA 11/81 Traffic Surveys by Roadside Interview (Clauses 6.1, 6.2, 6.5 and Figures 1 to 5 are superseded
by Ch. 8 of TSM) Scottish Addendum applicable for use in Scotland, Northern Ireland Addendum
applicable for use in Northern Ireland
- 94 -
TA 12/81 Traffic Signals on High Speed Roads Northern Ireland Addendum applicable for use in
Northern Ireland
TA 19/81 Reflectorisation of Traffic Signs (Clauses 7.6 and 7.7 are superseded by Ch.8 of TSM) Scottish
Addendum applicable for use in Scotland, Northern Ireland Addendum applicable for use in Northern
Ireland
TA 22/81 Vehicle Speed Measurement on All Purpose Roads
TA 23/81 Junctions and Accesses: Determination of Size of Roundabouts and Major/Minor Junctions
Scottish Addendum applicable for use in Scotland
TA 30/82 Choice between Options for Trunk Road Schemes Scottish Addendum applicable for use in
Scotland
TA 44/92 Capacities, Queues, Delays and Accidents at Road Junctions Computer Programs ARCADY/3
and PICADY/3 (TRRL)
TA 45/85 Treatment of Gaps in Central Reserve Safety Fences
TA 46/97 Traffic Flow Ranges for Use in the Assessment of New Rural Road Standards
TA 48/92 Layout of Grade Separated Junctions
TA 49/86 Appraisal of New and Replacement Lighting on Trunk Roads and Trunk Road Motorways [and
Amendment No 3 dated July 1990]Scottish Addendum applicable for use in Scotland
TA 56/87 Hazardous Cattle Crossings: Use of Flashing Amber Lamps Northern Ireland Addendum
applicable for use in Northern Ireland
TA 57/87 Roadside Features [Chapter 2 is superseded by TA 69/96] Scottish Addendum applicable for
use in Scotland, Northern Ireland Addendum applicable for use in Northern Ireland
TA 58/92 Traffic Signs and Road Markings for Lane Gains and Lane Drops on All Purpose Dual
Carriageway and Motorway Trunk Roads
TA 60/90 The Use of Variable Message Signs on All Purpose and Motorway Trunk Roads Northern
Ireland Addendum applicable for use in Northern Ireland
TA 61/94 Currency of the Traffic Signs Manual
TA 63/97 Convoy Working
TA 64/94 Narrow Lane and Tidal Flow Operations at Road Works on Motorways and Dual Carriage
Trunk Roads with Full Width hard Shoulders
TA 66/95 Police Observation Platforms on Motorways
- 95 -
- 96 -
HD 25/94 Foundations
HD 26/94 Pavement Design [Amendment No. 1 (March 1995, Amendment No. 2 ( February 1996) and
Amendment No. 3 (February 1998)]
HD 27/94 Pavement Construction Methods [Amendment No. 1 (March1995) and Amendment No.2
(February 1999)]
HD 28/94 Skidding Resistance [Amendment No. 1 (February 1999)]
HD 29/94 Structural Assessment Methods [Amendment No. 1 (November1996) and Amendment No.2
(May 1999)]
HD 30/99 Structural Assessment of Road Pavements
HD 31/94 Maintenance of Bituminous Roads [Amendment No. 1 (March1995) and Amendment No. 2
(February 1998)]
HD 32/94 Maintenance of Concrete Roads
HD 33/96 Surface and Sub-surface Drainage System for Highways
HD 34/93 Implementation and Use of the Quality Control Reporting System
HD 35/95 Technical Information
HD 36/99 Surfacing Material for New and Maintenance Construction
HD 37/99 Bituminous Surfacing Materials and Techniques [Amendment No.1 (May 1999)]
HD 38/97 Concrete Surfacing and Materials [Amendment No.1 (May 1999)]
TD 6/79 Transverse yellow bar markings at roundabouts
TD 7/80 Type Approval of Traffic Control Equipment (not applicable to Northern Ireland)
TD 9/93 Highway Link Design
TD 11/82 Use of Certain Departmental Standards in the Design and Assessment of Trunk Road Schemes
Scottish Addendum applicable for use in Scotland
TD 16/93 Geometric Design of Roundabouts
TD 17/85 Criteria for the Provision of Closed Circuit Television on Motorways Northern Ireland
Addendum applicable for use in Northern Ireland
TD 18/85 Criteria for the Use of Gantries for Traffic Signs and Matrix Traffic Signals on Trunk Roads
and Trunk Road Motorways. Scottish Addendum applicable for use in Scotland, Northern Ireland
Addendum applicable for use in Northern Ireland
- 97 -
TD 19/85 Safety Fences and Barriers [Amendment No. 1 dated 11/86]Scottish Addendum applicable for
use in Scotland
TD 22/92 Layout of Grade Separated Junctions
TD 23/86 Trunk Roads and Trunk Road Motorways: Inspection and Maintenance of Road Lighting
Scottish Addendum applicable for use in Scotland
TD 24/97 All Purpose Trunk Roads: Maintenance of Traffic Signals
TD 25/86 Trunk Roads and Trunk Road Motorways: Maintenance of Traffic Signs Scottish Addendum
applicable for use in Scotland, Northern Ireland Addendum applicable for use in Northern Ireland
TD 26/86 Trunk Roads and Trunk Road Motorways: Maintenance of Road Markings Northern Ireland
Addendum applicable for use in Northern Ireland
TD 27/96 Cross Sections and Headrooms [This document supersedes TD27/86 and SH 2/92]
TD 30/87 Design of Road Lighting for All Purpose Trunk Roads Scottish Addendum applicable for use in
Scotland, Northern Ireland Addendum applicable for use in Northern Ireland
TD 32/93 Wire Rope Safety Fence
TD 33/90 The Use of Variable Message Signs on All Purpose and Motorway Trunk Roads Northern
Ireland Addendum applicable for use in Northern Ireland
TD 34/91 Design of Road Lighting for Motorway Trunk Roads Scottish Addendum applicable for use in
Scotland, Northern Ireland Addendum applicable for use in Northern Ireland
TD 35/91 All Purpose Trunk Roads MOVA System of Traffic Control at Signals
TD 36/93 Subways for Pedestrians and Pedal Cyclists. Layout and Dimensions
TD 37/93 Scheme Assessment Reporting
TD 39/94 The Design of Major Interchanges
TD 40/94 Layout of Compact Grade Separated Junctions
TD 41/95 Vehicular Access to All Purpose Trunk Roads
TD 42/95 Geometric Design of Major/Minor Priority Junctions
TD 45/94 Motorway Incident Detection and Automatic Signalling (MIDAS)
TD 46/94 Motorway Signalling
TD 49/97 Mobile Lane Closures Supersedes those of Chapter 8, Topic 6 that deal with Mobile Lane
Closures
- 98 -
- 99 -
Asmussen (1984). International Calibration Study of traffic conflict techniques.NATO ASI Series F:
Computer and System Sciences. Springer-Verlag. Berlin, Heidelberg, New York, Tokyo.
Austroads (2000). Bicycles Guide to traffic engineering practice. Standards Australia, Part 14. Austroads,
Sydney, Australia.
Automobile Association (1996). Britains roads- a national asset? AA Public Affairs, Basingstoke.
Automobile Association (1999a). The Great British Motorist 2000.AA, Basingstoke.
Automobile Association (1999b). What goes wrong in Highway Design-and how to put it right. AA
Policy, TMS Consultancy. AA, Basingstoke.
Baguley C J and S D Ray (1989). Behavioural assessment of speed discrimination at traffic signals. TRL
Report RR 177. TRL Limited, Crowthorne.
Barker J (1997). Trials of Rural Safety Measures. TRL Report 202. TRL Limited, Crowthorne.
Barker J and R D Helliar-Symons (1996). Count-down signs and Roundel markings trials. TRL Report
TRL201. TRL Limited, Crowthorne.
Barker J, S Farmer and D Nicholls (1998). Injury accidents on rural single-carriageway roads, 1994-95:
An analysis of STATS19 data. TRL ReportTRL304. TRL Limited, Crowthorne.
Barker J, S Farmer and M Taylor (1999). The development of accident-remedial intervention levels for
rural roads. TRL Report TRL425. TRL Limited, Crowthorne.
Barker J, S Farmer and M Taylor (to be published 2001). Accident analysis methods for rural roads a
technical manual. TRL Report. TRL Limited, Crowthorne.
Baruya A (1998). Master: Speed accident relationships on different kinds of European roads. Master
deliverable D7, EU Commission, Brussels.
Binning J C (1998). Visual Oscady/4 user guide. TRL Application GuideAG25. TRL Limited,
Crowthorne.
Binning J C (2000a). Arcady 5 (UK) user guide. TRL Application GuideAG36. TRL Limited,
Crowthorne.
Binning J C (2000b). Visual Picady 4 (International) User Guide. TRL Application Guide AG 30. TRL
Limited, Crowthorne.
BINNING J C and M R CRABTREE (1999). TRANSYT 11 user guide. TRL Application Guide AG35.
TRL Limited, Crowthorne.
Boulter P (1999). Remote sensing of vehicle emissions near two traffic calming measures in Gloucester.
TRL Report 423. TRL Limited, Crowthorne.
- 100 -
Boulter P (2000). The impacts of the Safer City Project on road traffic emissions in Gloucester:
1996-1998. TRL Report 444. TRL Limited, Crowthorne.
Boulter P G, A J Hickman , S Latham and R Layfield (TRL Limited), PDavison and P Whiteman (AEA
Technology) (to be published 2001). The impact of traffic calming measures on vehicle exhaust
emissions. TRL Report482. TRL Limited, Crowthorne.
Boyle A J and C C Wright (1984). Accident migration after remedial treatment at accident blackspots.
Traffic Engineering and Control, 25 5.
Bretherton R D, K Wood and G T Bowen (1998). SCOOT Version 4. In:
Proceedings of 9th International Conference on Road Transport Information and Control. Institution of
Electrical Engineers, London.
Brindle R E (1986). The relationship between traffic management, speed and safety in neighbourhoods.
In: Proceedings of 13th ARRB Conference 13(a), pp90-102. Australian Research Board, Vermont South.
British Board of Agrment (1998). Guidelines document for the assessment and certification of
high-friction surfaces for highways. Final draft, SG1/98/075, 21 September 1998. British Board of
Agrment, Watford.
British Motorcyclists Federation (1997). Powered two-wheeler use into the twenty first century. British
Motorcyclists Federation.
Brown M (1995). The design of roundabouts. TRL State-of-the-Art Review. The Stationery Office,
London.
BrowN N C (1996). A Guide to the design and provision of secure parking for motorcycles. MAG, Kings
Norton, Birmingham.
Bryan-Brown K (1995). The Effects of a Childrens Traffic Club. In: Road Accidents Great Britain: 1994:
The Casualty Report. Department of the Environment, Transport and the Regions. The Stationery Office,
London.
Bryan-Brown K and G Harland (1999). The Childrens Traffic Club in Scotland. Scottish Executive
Research Unit Report. The Stationery Office, London.
BSI (1998). BS EN 1317-1, Road Restraint Systems-Part 1: Terminology and general criteria for test
methods. The British Standards Institution, London.
BSI (1998). BS EN 1317-2, Road Restraint Systems-Part 2: Performance classes, impact test acceptance
criteria and test methods for safety barriers. British Standards Institution, London.
BSI (1999). PrEN 1317-4: 17-18 June 1999 Not published at committee stage. Road Restraint Systems
Part 4: Barrier systems: Terminals and Transitions Performance classes, impact test acceptance criteria
and test methods. Not published at committee stage. British Standards Institution, London.
- 101 -
BSI (2000). BS EN 1317-3. Road Restraint Systems-Part 3: Crash Cushions Performance classes, impact
test acceptance criteria and test methods. British Standards Institution, London.
BSI (2000). PrEN 1317-6: September 2000. Part 6: Road restraint systems Pedestrian restraint systems
Part 6: Pedestrian Parapet. Not published at committee stage. British Standards Institution, London.
BSI (2000). PrEN 1317-5: June 2000 Not published at committee stage.Part5: Road Restraint
Systems-Product Requirements, Durability and evaluation of conformity. Not published at committee
stage. British Standards Institution, London.
BSI (2000). BS EN 12767. Passive safety of support structures for road equipment Performance
requirements and test methods. British Standards Institution, London.
BSI (1998). BS6579: (Parts 1 to 8). Safety fences and barriers for highways. British Standards Institution,
London.
BSI (1999). BS6779: (Parts 1 to 4). Highway parapets for bridges and other structures. British Standards
Institution, London.
Burrow I J (1987). Oscady: A computer program to model capacities, queues and delays at isolated traffic
signals. TRRL Report RR105. TRL Limited, Crowthorne.
Child Accident Prevention Trust (2000a). Accidents and child development. Child Accident Prevention
Trust.
Child Accident Prevention Trust (2000b). Checking child car seats. Child Accident Prevention Trust.
Clayton A, G Davies, G Harland and V Platt (1998). Road safety education in schools. In: Proceedings of
the road safety education conference, York. The Stationery Office, London.
Cloke J, D Webster, P Boulter, G Harris, R Stait, P Abbott and LChinn (1999). Traffic calming:
Environmental assessment of the Leigh Park Area Safety Scheme. TRL Report 397. TRL Limited,
Crowthorne.
Coolican H (1994). Research Methods in Psychology. Holder and Stoughton, London.
Corbett C and D Simon (1999). The effect of speed cameras: how drivers respond. Centre for Criminal
Justice Research. Br).
County Surveyors Society and department of Transport(1994a). Traffic calming in practice. Landor
publishing Ltd, London. unel University.
County Surveyors Society (2000). The Use of Coloured Road Surfaces. CSS (email:
CSS@derbyshire.gov.uk
County Surveyors Society and department of Transport(1994b). Village speed control (VISP) Working
Group final report. DTLR, London.
- 102 -
CTC Working for Cycling (2000a). Delivering Safer roads- Managing the interaction of cycles and
lorries. CTC.
CTC Working for Cycling (2000b). The Cycle Safe Framework- A handbook for delivering cycling
safely. CTC.
Danish Roads Directorate (1999). Speed management in urban areas. Report number 168. Dumas.
Davies D G, M C Taylor, M C Ryley and M E Halliday (1997). Cyclists at roundabouts the effects of
Continental design on predicted safety and capacity. TRL Report TRL285. TRL Limited, Crowthorne.
Davies D G, P Emmerson and G GardneR (1998). Achieving the aims of the National Cycling Strategy:
Summary of TRL research. TRL Report TRL365.TRL Limited, Crowthorne.
Delhomme P, V Truls et al. (1999). Evaluation of road safety media campaigns: an overview of 264
evaluated campaigns and some meta-analysis on accidents. Arceuil, INRETS, Paris.
Denning J H (1978). Epoxy-resin / calcined bauxite surface dressing on A1,Sandy, Bedfordshire:
skid-resistance measurements 1968 to 1977. Department of the Environment, Department of Transport
TRRL Report LR 867. TRL Limited, Crowthorne.
Denning J H, and J Carswell (1983). High-performance surface dressing:3. Properties of thermosetting
binders related to road performance. Department of the Environment, Department of Transport TRRL
Supplementary Report 798.TRL Limited, Crowthorne.
Department of the Environment, Transport and the Regions (1992).Residential roads and footpaths.
Layout Considerations, Second Edition. Design Bulletin 32. HMSO, LONDON.
Department of the Environment, Transport and the Regions (1993).Road Traffic Regulation Act 1984:
sections 81-85 Local speed limits: Circular Roads 1/93. The Stationery Office, London.
Department of the Environment, Transport and the Regions (1994).Traffic Signs Regulations and General
Directions. The Stationery Office, London.
Department of the Environment, Transport and the Regions (1996).The National Cycling Strategy. The
Stationery Office, London.
Department of the Environment, Transport and the Regions (1998a).Trunk roads review: detrunking of
non-core routes, DTLR Roads Policy Division, Zone 3/01, Great Minster House, 76 Marsham Street,
London SW1P 4DR.
Department of the Environment, Transport and the Regions (1998b). Manual of Contract Documents for
Highway Works, Volume 1Specification for Highway Works, Series 1200 Traffic Signs. Highways
Agency, Scottish Office Development Department, Welsh Office, Department for the Environment for
Northern Ireland. The Stationery Office, London.
Department of the Environment, Transport and the Regions (1998c).Places, streets and movement. A
companion guide to Design Bulletin 32.Residential roads and footpaths. The Stationery Office, London.
- 103 -
Department of the Environment, Transport and the Regions (1998d). A new deal for trunk roads in
England. The Stationery Office, London. Department of the Environment, Transport and the Regions
(1998e).A new deal for trunk roads in England: Guidance on the new approach to appraisal. The
Stationery Office, London.
Department of the Environment, Transport and the Regions (1998f).Lets Decide Walk Wise. DTLR,.
Department of the Environment, Transport and the Regions (1998g). Kerbcraft smart strategies for
pedestrian safety. DTLR
Department of the Environment, Transport and the Regions (1999a).School Travel Strategies and Plans: A
Best Practice Guide for Local Authorities. Department of the Environment, Transport and the Regions
Department of the Environment, Transport and the Regions (1999b). The Highway Code. The Stationery
Office, London.
Department of the Environment, Transport and the Regions (1999c).20 mph speed limits. Circular 05/99.
The Stationery Office, London.
Department of the Environment, Transport and the Regions (1999d). Guidance on the Use of Tactile
Paving Surfaces. DTLR free literature.
Department of the Environment, Transport and the Regions REGIONS (1999e).Speed camera funding
eight pilot schemes announced. Press Notice 1226.
Department of the Environment, Transport and the Regions (2000a).Road accidents Great Britain: 1999
the casualty report. The Stationery Office, London.
Department of the Environment, Transport and the Regions (2000b). Tomorrows Roads Safer for
Everyone. DTLR Free Literature.
Department of the Environment, Transport and the Regions (2000c).Encouraging walking: advice to local
authorities. Available from DTLR, Charging and Local Transport Division, Zone 3/25 Great Minster
House, 76 Marsham Street, London SW1P 4DR Tel: 020 7944 2945, fax: 020 7944 2211, email:
cycle_walk@dft.gsi.gov.uk.
Department of the Environment, Transport and the Regions 2000d). New Directions in Speed
Management A Review of Policy. DTLR Free Literature.
Department of the Environment, Transport and the Regions (2000e).Guidance on full Local Transport
Plans. DTLR Free Literature.
Department of the Environment, Transport and the Regions (2000f).A good practice guide for the
development of Local Transport Plans. DTLR Free Literature.
Department of the Environment, Transport and the Regions (2000g). STATS20 Instructions for the
completion of road accident reports. The Stationery Office, London.
- 104 -
Department of the Environment, Transport and the Regions (2000h).Transport statistics Great Britain
2000 edition. The Stationery Office, London.
Department of the Environment, Transport and the Regions (2000i).Transport 2010: The Ten Year Plan.
http://www.dft.gov.uk\strategy
Department of the Environment, Transport and the Regions (2000j). Best value road condition surveys for
local highway authorities. Available from the DfT website.
Department of the Environment, Transport and the Regions (2000k)Best value Performance Indicators
2001/02.
Department of the Environment, Transport and the Regions (to be published 2001a). Accidents on rural
roads. Traffic Advisory Leaflet.
Department of the Environment, Transport and the Regions (to be published 2001b). Gloucester Safer
City Report.
Department of the Environment, Transport and the Regions (2001c). School travel.
http://www.dft.gov.uk/local-transport/schooltravel
Devon County Council (2000). Speed Management Strategy. Road Safety Group, Devon County Council.
Downing C S (1985). Pedal cycling accidents in Great Britain. In: Proceedings of Ways to Safer Cycling
at Institution of Civil Engineers. Department of Transport. London.
ETSC: European Transport Safety Council (1999). Safety of pedestrians and cyclists in urban areas.
ETSC, Brussels.
Farmer S, J Barker and N Mayhew (1998). A trial in Norfolk of interactive speed limit signs. Traffic
Engineering and Control 39 5.
FEMA (2000). Final Report of the Motorcyclists and Crash Barriers Project. FEMA-Federation of
Motorcyclists Associations.
Finch D J, P Kompfner, C R Lockwood and G Maycock (1994). Speed, speed limits and accidents. TRL
Report 58. TRL Limited, Crowthorne.
Forrester A and G Harland (2000). The effects of training children in Sunderland to be safer pedestrians.
To be published by TRL Limited.
GARDNER G and S M GRAY (1998). A preliminary review of rural cycling. TRL Report 310. TRL
Limited, Crowthorne.
Gloucester City Council and Gloucestershire County Council(1997). Gloucester Safer City road safety
strategy. Gloucester Safer City.
- 105 -
Government of Ireland (1996). A guide to road safety engineering in Ireland. The Stationery Office,
Dublin.
Grayson G B (1996). Behaviour adaptation: a review of the literature. TRL Report 254. TRL Limited,
Crowthorne.
Graphpad Software INC. (1999). The Gaussian distribution.
http://graphpad.com/articles/interpret/principles/gaussian.htm
Greater Manchester Passenger Transport Executive (1999).Quality Bus Routes consultation brochure,
Manchester.
Hall R D (1986). Accidents at four-arm single carriageway urban traffic signals. Southampton University.
TRRL Contractor Report CR65. TRL Limited, Crowthorne.
Harris G J, R E Stait, P G Abbott and G R Watts (1999). Traffic calming: vehicle generated noise and
ground-borne vibration alongside sinusoidal round-topa nd flat-top road humps. TRL Report 416. TRL
Limited, Crowthorne.
Hauer E (1992). How to do a before-and-after study: estimating the effect of highway and traffic
engineering measures on road safety. Lecture notes: Transport Canada Workshop, Ottawa, Canada, 1992.
Hauer E, and P Byer (1983). Bias-by-selection: the accuracy of an unbiased estimator. Accident Analysis
and Prevention 15 5.
Haynes J P, G Copley, S Farmer and R D Helliar-Symons (1993).Yellow bar markings on motorway slip
roads. TRL Report PR49. TRL Limited, Crowthorne.
Health and Safety Executive (2000). Crossing high-speed roads on foot during temporary traffic
management works. Construction Information Sheet No. 53. HSE Books, HSE, London.
Heath M (1995). Legal aspects of safety audit. Road safety auditors forum, Mott McDonald. 1995.
Helliar-Symons R D (1981). Yellow bar experimental carriageway markings accident study. TRL Report
LR 1010. TRL Limited, Crowthorne.
Helliar-Symons R D and N BUTLER (1995). M1 Chevron Trial accident study TRL Report PR118. TRL
Limited, Crowthorne.
Helliar-Symons R D, P Webster and A Skinner (1995a). The M1Chevron Trial. Traffic Engineering and
Control, 36 10.
Helliar-Symons R D, N Butler and P Critchell (1995b). Riblines in Hampshire. Traffic Engineering and
Control 36 7/8.
Higginson M (1999). Network Management Notes: Bus Priority. Oscar Faber.
- 106 -
Highways Agency (1996). The COBA Manual. Design Manual for Roads and Bridges. Volume 13,
Section 1, Part2, September 1996. The Stationery Office, London.
Highways Agency (1998). Towards a balanced transport system: The Highways Agencys Toolkit.
Highways Agency (2000a). Making the network safer Highways Agency strategic plan for safety.
Highways Agency.
Highways Agency (2000b). Encouraging Sustainable Travel. Highways Agency.
Highways Agency (2000c). Making Maintenance the Priority. Highways Agency.
Highways Agency (2000d). Trunk road traffic calming. Interim Advice Note28/00. Highways Agency.
HMSO (1985). Traffic Signs Manual Chapter 5: Road Markings. The Stationery Office, London.
HMSO (1986a). Traffic Signs Manual Chapter 3: Regulatory Signs. The Stationery Office, London.
HMSO (1986b). Traffic Signs Manual Chapter 4: Warning Signs. The Stationery Office, London.
HMSO (1986c). Shared Use by Cyclists and Pedestrians. Local Transport Note2/86. The Stationery
Office, London.
HMSO (1991). New Roads and Street Works Act. Section 26 Code of Good Practice, safety at street
works and road works. The Stationery Office, London.(2nd edition due 2001).
HMSO (1994a). Contract (Design and Management) Regulations 1994.The Stationery Office, London.
HMSO (1994b). The design and use of directional informatory signs. Local Transport Note 1/94. The
Stationery Office, London.
HMSO (1994c). The Traffic Signs Directions and General Directions. The Stationery Office, London.
HMSO (1995). The assessment and design of pedestrian crossings. Local Transport Notes 1/95 and 2/95.
The Stationery Office, London.
HMSO (1997). Traffic Signs Manual Chapter 7: The design of traffic signs. The Stationery Office,
London.
HMSO (1998a). Manual of contract documents for Highway Works. Volume 1 Specification for Highway
Works. The Stationery Office, London.
HMSO (1998b). Manual of contract documents for Highway Works. Volume 3 Highway Construction
Details. The Stationery Office, London.
HMSO (1998c). Traffic Signs Manual Chapter 8: Traffic safety measures and signs for road works and
temporary situations. Consolidated issue. The Stationery Office, London.
- 107 -
HMSO (1998d). The Installation of Traffic Signals and Associated Equipment. Local Transport Note
1/98. The Stationery Office, London.
Hoinville J, R Jowell and Associates (1978). Survey Research Practice. Heinemann Educational Books.
London.
Hooke A, J Knox and D Portas (1996). Cost benefit analysis of traffic light and speed cameras. Police
Research Series Paper 20. Home Office: Police Research Group.
Hughes W and G Amis (1996). Accidents on rural single-carriageway A class roads. Cambridgeshire CC
and AA Foundation for Road Safety Research.
Hughes W, G Amis and A Walford (1997). Accidents on rural roads: dual-carriageway A class roads.
Cambridgeshire County Council and AA Foundation for Road Safety Research.
Hunt J G, J D Griffiths and T S Moses (1990). An evaluation of the effect of SCOOT on road safety. TRL
Contractor Report CR 236. TRL Limited, Crowthorne.
Institution of Highways and Transportation (1990a). Guidelines forUrban Safety Management. IHT,
London.
Institution of Highways and Transportation (1990b). Accident Reduction and Prevention. IHT, London.
Institution of Highways and Transportation (1996a), Department of Transport, Bicycle Association, and
Cyclists Touring Club. Cycle-Friendly Infrastructure. Guidelines for Planning and Design. IHT, London.
Institution of Highways and Transportation (1996b). Guidelines for Safety Audit of Highways. IHT,
London.
Institution of Highways and Transportation (1997). Transport in the Urban Environment. IHT, London.
Institution of Highways and Transportation (1998). Cycle Audit and Cycle Review. IHT Guidelines. IHT,
London.
Institution of Highways and Transportation (1999a). Guidelines for providing for journeys on foot. IHT,
London.
Institution of Highways and Transportation (1999b). Planning for public transport in developments. IHT,
London.
Institution of Highways and Transportation (1999c). Guidelines for Rural Safety Management. IHT,
London.
Institution of Highways and Transportation (1999d). Reducing Mobility Handicaps: towards a
Barrier-Free Environment. IHT, London.
Institution of Highways and Transportation (2000). Parking problems hospitals face. IHT, London.
- 108 -
Institution of Highways and Transportation (2000). Providing for journeys on foot. IHT, London.
Institution of Highways and Transportation (to be published 2001).Environmental management of
highways guidelines on good practice. IHT, London.
Institution of Highways and Transportation (to be published 2001).Pedestrianised High Streets planning,
design and management. IHT, London. James H (1991). Under-reporting of road traffic accidents. Traffic
Engineering and Control. 32 12.
Kallberg V-P, R E Allsop, H Ward, R Van Der Horst and A Varhelyi(1998). Recommendations for speed
management strategies and policies. Deliverable D12 of the MASTER (Managing Speeds of Traffic on
European Roads) Project. VTT Communities and infrastructure, Finland.
http://www.vtt.fi/yki/yki6/master/master.htm
Kennedy J V, R D Hall and S R Barnard (1998). Accidents at urban mini-roundabouts. TRL Report TRL
281 TRL Limited, Crowthorne.
Kulmala R (1994). Measuring the safety effect of road measures at junctions. Accident Analysis and
Prevention. 26 6.
Lawson S (1991). Red light running: accidents and surveillance cameras. AA Foundation for Road Safety
Research and Birmingham City Council.
Layfield R (2000). Home zones: monitoring the DTLR pilot programme. Proceedings of Conference on
Reducing Traffic Impacts on Local Communities. Aston University. PA2626-00. TRL Limited,
Crowthorne.
Layfield R E and D I Parry (1998). Traffic calming Speed cushion schemes. TRL Report 312. TRL
Limited, Crowthorne.
Layfield R E and G Maycock (1986). Pedal cyclists at roundabouts. Traffic Engineering and Control,
1986.
Local Authorities Association (1989). Highway Maintenance Code of Good Practice.
Local Authorities Association (1996). Road Safety Code of Good Practice.
London Research Centre (1999a). Levels of accident risk in Greater London. London Research Centre.
London Research Centre (1999b). London Accident Analysis Unit Annual Report 1998. London Research
Centre.
London Research Centre (1999c). Towards the Year 2000: monitoring casualties in Greater London.
London Research Centre.
Mackie A M, H A Ward and R T Walker (1990). Urban safety project 3:Overall evaluation of area-wide
schemes. TRL Report RR263. TRL Limited, Crowthorne.
- 109 -
MACKIE A (1998). Urban speed management methods. TRL Report TRL 363.TRL Limited, Crowthorne.
Maher M J and L J Mountain (1988). The identification of accident blackspots: a comparison of current
methods. Accident Analysis and Prevention. 20 2.
MARTENS M and N KAPTEIN (1988). Speed behaviour before and after road design modifications a
meta review. Deliverable 2 of the MASTER (Managing Speeds of Traffic on European Roads) Project.
VTT Communities and infrastructure, Finland.
http://www.vtt.fi/yki/yki6/master/master.html
Maycock G and R D Hall (1984). Accidents at 4-arm roundabouts. TRL Report LR1102. TRL Limited,
Crowthorne.
Maycock G, P J Brocklebank and R D Hall (1998). Road layout design standards and driver behaviour.
TRL Report TRL332. TRL Limited, Crowthorne.
Mccoll E and R Thomas (2000). Design and administration of postal self-completion surveys. National
Centre for Social Research. Economic and Social Research Council (ESRC).
Medd I (2000). The use of coloured road surfaces CSS Accident Reduction Working Group Guidelines.
Proceedings of CSS Conference Safety through engineering, CSS, October 2000, Leamington Spa.
Mills Paula J (1989). Pedal cycle accidents a hospital based study. TRL Research Report RR220. TRL
Limited, Crowthorne.
Morgan J (1997). Roundabouts in Continental Europe designed with cycle facilities of cycle-thinking.
TRL Report 302. TRL Limited, Crowthorne.
Mountain L and B FAWAZ (1992). The effects of engineering measures on safety at adjacent sites.
Traffic Engineering and Control, 33, 1.
Mountain L, B Fawaz and L Sineng (1992). The assessment of changes in accident frequencies on link
segments: a comparison of four methods. Traffic Engineering and Control, 33, 7.
Mountain L (1998). Regression, migration and trend: road accident reductions arent always what they
seem.
Muskaug R (1993). Drivers acceptance of automatic traffic surveillance. Traffic Engineering and Control,
34 4.
National Assembly for Wales (2000). Traffic calming trials: a review of traffic calming schemes on the
trunk and principal road network in Wales. Highways Directorate, Colwyn Bay.
Nicholls J C (1997). Laboratory Tests on High-Friction Surfaces for Highways. TRL Report 176. TRL
Limited, Crowthorne.
Nicholls J C (1998). Trials of High-Friction Surfaces for Highways. TRL Report 125. TRL Limited,
Crowthorne.
- 110 -
Pacts-Parliamentary Advisory Council for Transport Safety(1996). Taking action on Speeding. Pacts.
Pacts-Parliamentary Advisory Council for Transport Safety(1999). Road traffic law and enforcement: a
driving force for casualty reduction. Pacts.
Persaud B (1987). Migration of accident risk after remedial treatment at accident blackspots. Traffic
Engineering and Control, 28 1.
Pickering D, R D Hall and M Grimmer (1986). Accidents at rural T-junctions. TRL Report RR65. TRL
Limited, Crowthorne.
Ross Silcock (1998). Pedestrian behaviour and exposure to risk final report and appendices. ISBN
0-9533722-2-7. Babtie Ross Silcock. Newcastle-upon-Tyne.
Royal Society for the Prevention of Accidents (1990). Handbook for road safety officers. RoSPA,
Birmingham.
Royal Society for the Prevention of Accidents (1995a). Road Safety: A Managers Guide. RoSPA,
Birmingham.
Royal Society for the Prevention of Accidents (1995b). Road Safety Engineering Manual. RoSPA,
Birmingham.
Royal Society for the Prevention of Accidents (2000). Guidelines for the management and operation of
practical cyclist training schemes. RoSPA, Birmingham.
Ryley T J (1996). Advanced stop lines for cyclists: the role of central cycle laneapproaches and signal
timings. TRL Report 181. TRL Limited, Crowthorne.
Savill T and K Bryan-Brown (1996). The effectiveness of cycle training schemes. TRL Report 214. TRL
Limited, Crowthorne.
Sayer I A and D I Parry (1994). UK Speed control using chicanes a trial at TRL. TRL Project Report 102.
TRL Limited, Crowthorne.
Sayer I A, D I Parry and J K Barker (1998). Traffic calming An assessment of selected on-road chicane
schemes. TRL Report 313. TRL Limited, Crowthorne.
Sayer I A, D A Nicholls and R E Layfield (1999). Traffic calming: Passenger and rider discomfort at
sinusoidal, round-top and flat-top humps a track trial at TRL. TRL Report 417. TRL Limited, Crowthorne.
Scottish Executive (1999). Cycling by design.
Scottish Office (1995a). Running the red an evaluation of Strathclyde polices red light camera initiative.
Research Findings No 7. The Scottish Office Central Research Unit. Edinburgh.
Scottish Office (1995b). The Scottish Office Road Safety Plan.HMSO Scotland.
- 111 -
Scottish Office (1999). 20 mph speed limits. SODD Circular No 13/99.Scottish Office.
Scottish Office Industry Department Roads Directorate (1994).Hazard warning signs and markings on
bends on trunk roads. Roads Directorate of the Scottish Office.
Silcock D, K Smith, D Knox and K Beuret (1999). What limits speed? Factors that affect how fast we
drive. AA Foundation of Road Safety Research, Basingstoke.
Statsoft Inc. (1995). Experimental Design (Industrial DOE).
http://www.statsoftinc.com/textbook/stexdes.html
Stewart D (1995). Legal aspects of safety audit. Road safety auditors forum, Mott McDonald, 1995.
Suffolk County Council (1997). 30mph speed limit initiative: speed monitoring. Report 960297/R01.0.
Suffolk County Council, Suffolk.
Sustrans and OVE ARUP (1997). The national cycle network guidelines and practical details. Sustrans
Limited.
TAS Partnership (2000). Quality Bus Infrastructure: A Manual and Guide. The Tas Partnership Limited.
Taylor D and M Tight (1996). Feet First: Public Attitudes and consultation in traffic calming schemes. A
report for Transport 2000.
Taylor M C and J K Barker (1992). Injury accidents on rural single-carriage wayroads an analysis of
STATS 19 data. TRL Report RR365.TRL Limited, Crowthorne.
Taylor M C and A H Wheeler (2000). Accident reduction resulting from village traffic calming. European
Transport Conference, Cambridge. Proc. Seminar J. Assoc. for European Transport.
Taylor M C, R D Hall and K Chatterjee (1996). Accidents at 3-arm traffic signals on urban
single-carriageway roads. TRL Report 135. TRL Limited, Crowthorne.
Taylor M C, Lynam D A and Baruya A (2000). The effects of drivers speed on the frequency of road
accidents. TRL Report 421. TRL Limited, Crowthorne.
Taylor M C, A Mackie and A Baruya (to be published 2001). Vehicle speed distributions and safety on
urban roads. TRL Report. TRL Limited, Crowthorne.
Thames Valley Police (1999). Speed limits: The facts. Thames Valley Police.
The Motorcycle Industry Association (1999). Powered Two Wheelers The SMART Choice in Local
Transport Plans. Motor Cycle Industry Association Ltd.
Thomson J A, A K Tolmie, H C Foot and B Mclaren (1996). Child development and the aims of road
safety education: a review and analysis. The Stationery Office, London.
- 112 -
Transport and Road Research Laboratory (Now TRL Limited) (1987).The Traffic Conflict Technique
Guidelines. IHT, London.
Transport for London Street Management (2000). Accidents andcasualties on Londons roads 1999.
Transport for London Street Management.
Trimpop R M, and G J S Wilde (1994). Challenges to accident prevention: the issue of risk compensation
behaviour. STYX Publications, Postbus 2659,9704 CR Groningen, The Netherlands.
TRL Limited (1999). SafeNET User Guide v1. TRL Application Guide AG34. TRL Limited, Crowthorne.
Urban Design Alliance (2000). Returning roads to residents. ICE, London.
Vincent R A and J R Peirce (1988). MOVA: traffic responsive, self-optimising signal control for isolated
intersections. TRRL Research Report RR 170. TRL Limited, Crowthorne.
Vincent R A, J R Peirce and P J Webb (1991). MOVA traffic control manual. TRL Application Guide
AG10. TRL Limited, Crowthorne.
Walmsley D A and I Summersgill (1998). The relationship between road layout and accidents on modern
rural trunk roads. TRL Report TRL334.TRL Limited, Crowthorne.
Warwickshire County Council (1997). A safety auditors view of roundabout design. West Midlands
Region Road Safety Audit Forum An Update. Warwick.
Watts G R (1973). Road humps for the control of vehicle speeds. TRRL Laboratory Report LR597. TRL
Limited, Crowthorne.
Watts G R, and G J Harris (1996). Traffic calming: Vehicle Generated Ground-borne Vibration alongside
Speed Control Cushions and Road Humps. TRL Report 235. TRL Limited, Crowthorne.
Watts G R, G J Harris and R E Layfield (1997). Traffic calming: vehicle generated ground-borne
vibration alongside speed control cushions and roadhumps. TRL Report 235. TRL Limited, Crowthorne.
Webb P J and S Harrison (1992). The safety record of MOVA signal controlled junctions. TRL
Unpublished Working Paper WP/TO/102. TRL Limited, Crowthorne.
Webster D C (1993a). Grounding of vehicles on road humps. Traffic Engineering and Control 34 07/08.
Webster D C (1993b). Road humps for controlling vehicle speeds. TRL Project Report PR 18. TRL
Limited, Crowthorne.
Webster D C (1994). Speeds at thumps and low height road humps. TRL Project Report PR 101. TRL
Limited, Crowthorne.
Webster D C (1995). Traffic calming vehicle-activated speed limit reminder signs. TRL Report 177. TRL
Limited, Crowthorne.
- 113 -
Webster D C (1998). Traffic calming Public attitude studies: a literature review. TRL Report 311. TRL
Limited, Crowthorne.
Webster D C (2000). Traffic calming a literature search on the design and performance of traffic calming
measures. TRL Report TRL439. TRL Limited, Crowthorne.
Webster D C and R E Layfield (1993). An assessment of rumble strips and rumble areas. TRL Report
PR33. TRL Limited, Crowthorne.
Webster D C and R E Layfield (1996). Traffic calming Road hump schemes using 75 mm high humps.
TRL Report 186. TRL Limited, Crowthorne.
Webster D C and R E Layfield (1998). Traffic calming Sinusoidal, H and S humps. TRL Report 377. TRL
Limited, Crowthorne.
Webster D C and A M Mackie (1996). Review of traffic calming schemes in20 mph zones. TRL Report
215. TRL Limited, Crowthorne.
Webster D C and P A WELLS (2000). The characteristics of speeders. TRL Report TRL 440. TRL
Limited, Crowthorne.
Webster P B, A J Skinner and R Helliar-Symons (1992). Installation and evaluation of chevron markings
on motorways. TRL Report CR304. TRL Limited, Crowthorne.
Welsh Office (1999). 20 mph speed limits. Circular 28/99. Welsh Office.
Wheeler A (1995). Advanced stop-lines for cyclists a simplified layout. Traffic Engineering and Control
36 5.
Wheeler A and M Taylor (1995). Reducing speeds in villages: the VISP study. Traffic Engineering and
Control, 36 4.
Wheeler A and M C Taylor (1999). Traffic Calming on Major Roads: Final Report. TRL Report 385. TRL
Limited, Crowthorne.
Wheeler A H and M C Taylor (2000). Changes in accident frequency following the introduction of traffic
calming in villages. TRL Report TRL452. TRL Limited, Crowthorne.
Wheeler A, G Harris, L Chinn, M Taylor and P G Abbott (1997). A traffic calming scheme at Costessey,
Norfolk. TRL Report 364. TRL Limited, Crowthorne.
Wheeler A, P G Abbott, D J Lawrence and S M Phillips (1996). Traffic calming on major roads: the A49
trunk road at Craven Arms, Shropshire. TRL Report 212. TRL Limited, Crowthorne.
Wheeler A, P G Abbott, N S Godfrey, S M Phillips and R Stait (1997).Traffic calming on major roads: the
A47 trunk road at Thorney, Cambridgeshire.TRL Report 238. TRL Limited, Crowthorne.
- 114 -
Wheeler A, M C Taylor and J Barker (1994). Speed Reductions in 24villages: details from the VISP
study. Project Report PR85. TRL Limited, Crowthorne.
Wilde G J S (1994). Target risk. PDE Publications. Toronto.
Winn R. (1995). Running the red: an evaluation of Strathclyde Polices red light camera initiative. The
Scottish Office Central Research Unit. HMSO, Edinburgh.
Winnett M (1994). A review of speed camera operations in the UK. Traffic management and road safety.
Proceedings of 22nd PTRC European Transport Forum, Seminar J.
Winnett M (TRL), E Woodgate (TRL) and N Mayhew (Norfolk County council) (1999). Interactive Fibre
Optic Signing at a Rural Crossroad. TRL Report TRL 401. TRL Limited, Crowthorne.
Winnett M A, E Woodgate and A Mackie (1997). Rural Safety Measures Vehicle Activated Bend Sign.
TRL Project Report PR/TT/037/97. (Unpublished Report). TRL Limited, Crowthorne.
Wright C C and A J Boyle (1987). Road accident causation and engineering treatment: a review of current
issues. Traffic Engineering and Control, 28 9.
Wright C C, C R Abbess and D F Jarrett (1986). Estimating the regression-to-mean effect associated with
road accident blackspot treatment: towards a more realistic approach. Accident Analysis and Prevention.
20 3.
7. Abbreviations
AA
Automobile Association
AIP
ARCADY
BITER
CAPT
CCTV
CPRE
CSS
DTLR
DoE NI
DMRB
85th
percentile
The speed at or below which 85 per cent of the vehicles in aspeed speed measurement
sample set were travelling
- 115 -
ETP
FYRR
GB
Great Britain
GIS
HA
Highways Agency
HGV
HMSO
HSE
ICE
IHT
IRSO
KSI
(casualty)
KSI
(accident)
LAA
LGV
LHA
LTN
LTP
MCAP
MOLASSES Monitoring Local Authority Safety Schemes A database managed by TRL Limited
storing information about local and trunk road safety schemes which can be interrogated
to ascertain the overall effectiveness of specific engineering measures. (See also the web
site www.trl.co.uk/molasses)
MOVA
NATA
NPV
NVQ
OSCADY
- 116 -
PIA
PICADY
PSV
PTRC
PVR
Per vehicle record used to describe the type of automatic equipment that can measure
speeds of all individual vehicles
RAGB
RoSPA
RSE
RSO
RSR
RTRA
SafeNET
TRL software used to estimate the frequency of accidents on a network (when traffic and
pedestrian flow and geometric information are provided)
SCOOT
SO
Scottish Office
STATS19
Database of standardised accident reports (using theSTATS19 report form) sent to the
Department of the Environment, Transport and the Regions by all police forces in Great
Britain
STATS20
Document with instructions for the completion of STATS19 road accident reports
TAL
TPP
TRANSYT
TRL
TSM
TSO
TSRGD
TWMV
UK
United Kingdom
USM
- 117 -
UTC
VISP
WO
Welsh Office
8. Feedback
We expect to review and update this guide. If you would like to make any suggestions or can offer any
further site examples, please e-mail Miss C Britt:Caroline.Britt@dft.gsi.gov.uk, in the first instance, by 31
December 2001.We will consider all feedback in preparing the next update.
Appendix A
Road accident countermeasures
A.1
Anti-skid/high-friction surfacings
High-friction surfacings: suburban
High-friction surfacings: urban
A.2
A.3
A.4
Speed cameras
Speed cameras: various urban locations
Speed camera: suburban
A.5
Chevron markings
Chevrons: motorway
A.6
Chicanes/narrowings
Chicanes: residential estate
Chicanes: major road traffic calming
A.7
A.8
Cycling facilities
Cycle track at roundabout: use of coloured road surfacing
Annular cycle track at multiple roundabout
- 118 -
A.9
Gateways
Gateways: rural village
Gateways and other treatments: rural village
- 119 -
Appendix B
Statistical Tests
B.1 Students t-test for comparison of samples
Table of t-distribution
B.2 Kolmogorov-Smirnov test
Table of 2
B.3 The Tanner k test
B.4 The Chi-Squared test
B.5 Test for statistical significance between two proportions
B.6 Regression-to-the-mean correction
- 120 -